Bridging the Physical & Digital Worlds

Ahmad Emranian — Spatial Computing & Real-Time R&D

Developing Virtual Production and 3D Reconstruction solutions through R&D and the use of real-time 3D and AI technologies.

Abu Ecosystem Products

• Abu Clock — Software & Hardware-Ready Engine / SMPTE Timecode, Timestep (Hz) & PTP/gPTP Master Clock

  Cross-platform software and hardware-ready timing engine engineered to deliver broadcast-grade SMPTE ST 2059-1 / ST 12-1 timecode (HH:MM:SS:FF.sf) and deterministic frequency timestep (Hz) synchronization across standard local area networks (Gigabit Ethernet, USB-C LAN, and Wi-Fi). Designed as the temporal backbone of the Abu ecosystem alongside Abu Track & Scan and Abu Cam, it bridges Virtual Production (VP), AR broadcast, and 3D Gaussian Splatting (3DGS) spatial capture pipelines with enterprise broadcast timing.

  Specifications: Cross-Platform Software Suite: Windows 10/11 (x86_64/ARM64), macOS 13+ (Apple Silicon/Intel), Linux (x86_64/ARM64), iOS 16+ & Android 10+ via Tauri v2 Digital Slate GUI | User-Space OverlayClock Virtual PLL & Two-State Kalman Servo with Hybrid Spinlock Timestep Ticker (< 5 μs software jitter, zero root/admin required) | Supported Cadences: 23.976, 24, 25, 29.97 (DF/NDF), 30, 48, 50, 59.94 (DF/NDF), 60, 120 fps & LiDAR (10, 15, 20 Hz) + Scheduled Future-Time Capture Triggers (T_now + 50 ms) | Optional Phase 3 Dedicated Hardware Grandmaster Appliance: Raspberry Pi CM4 / Pi 5 (/dev/ptp0), Maxim DS3231 TCXO (±2 ppm holdover) & optional Intel I210-T1 PCIe NIC (/dev/ptp1)

  Protocols: IEEE 1588-2019 (PTPv2.1), SMPTE ST 2059-1/2 (SM TLV 0x000E9B), IEEE 802.1AS (gPTP), Client-Server PTP (CSPTP) & SNTPv4 | AbuSync Unprivileged CSPTP over UDP (Ports 50319 / 50320 + Zero-Config mDNS _abuclock._udp.local.) | SMPTE RIS OpenTrackIO (OTrk JSON/CBOR) & Extended FreeD-TC (0xD1 / 0xD2) Deterministic Camera Telemetry | Mozilla UniFFI SDK (libabuclock) with Native Swift, C/C++ ABI, C# & Python Bindings + .abuscan.json / COLMAP Session Manifests

  Engines: Unreal Engine 5 via Native AbuClockUE Plugin (UAbuClockTimecodeProvider, UAbuClockCustomTimeStep & FAbuTrackLiveLinkSource) | NVIDIA Omniverse via Native omni.abu.clock Kit Extension (Synchronized omni.timeline & OpenUSD Stages) | Abu Track & Scan (AbuClockKit iOS Optical Exposure Midpoint Wrapper & AbuClockWinClient Windows C++ Aggregator) | RoboSense E1R & XGRIDS L2PRO LiDAR Sensors + RealityCapture, COLMAP, LixelCyberColor & Jawset Postshot

• Abu Track & Scan — Software / Spatial Telemetry, LiDAR Scanning & VP Ecosystem

  End-to-end 60Hz spatial tracking, real-time LiDAR point cloud reconstruction, and sub-millisecond camera synchronization suite engineered for Virtual Production (Unreal Engine 5), 3D Gaussian Splatting (Postshot), and High-Precision Photogrammetry (Agisoft Metashape & RealityScan)—combining an iOS App, Windows Workstation Suite, and Native UE5 Live Link C++ Plugin.

  Specifications: iOS App (iPhone 13–16 Pro/Max & iPad Pro LiDAR): 60Hz ARKit VIO + LiDAR Depth, 4K H.264/HEVC RTP Video Streamer, Multi-Lens & 180° Parallel Selfie Rig Solver, NTP Sub-ms Clock Sync | Windows Workstation Suite: 60FPS Interactive 3D Viewport & AR Comp, Time-Shift Jitter Playout Buffer, PyAV 4K Hardware Decoder, Zero-Copy GPU Spout2 Sender & OpenCV Chessboard Lens Calibrator | DSLR / Mirrorless Hot-Shoe Rig Mode: Sub-Second EXIF Timestamp Matching, Quaternion SLERP Pose Interpolation & 3D Lever-Arm Offsets for Sony, Canon, Nikon & RED | Live Voxel-Filtered LiDAR Mapping (3cm Grid) with Statistical Outlier Removal (SOR) & 4K WebRTC Pixel Streaming Return Monitor

  Protocols: Real-Time 6-DoF UDP Telemetry: AbuTrack JSON, FreeD (D1 29-byte Binary), OpenTrackIO & Raw CSV | 4K Ultra-HD Hardware Video Streaming: H.264 (AVC) & H.265 (HEVC) over RTP/UDP (RFC 6184 / RFC 7798) + Zero-Copy Spout2 GPU Texture Sharing | Sub-Millisecond NTP Clock Synchronization & Portable Optical Lens Profiles (.atlp with Focal Length, Principal Point & K1..K4, T1..T2 Distortion) | One-Click 3DGS & Photogrammetry Exports: Jawset Postshot COLMAP Rigs, Metashape & RealityScan Georeferenced CSVs, Embedded EXIF GPS & Clean .ply Point Clouds

  Engines: Unreal Engine 5.x via Native C++ Plugin (LiveLinkAbuTrack — ULiveLinkCameraRole, Live Focal Length & K1/K2 Lens Distortion Blueprint Metadata) | Jawset Postshot (3D Gaussian Splatting Training from 60Hz Pose Logs & LiDAR Point Clouds) | Agisoft Metashape & Epic Games RealityScan / RealityCapture | OBS Studio, TouchDesigner & Broadcast Switchers via Zero-Copy Spout2 GPU Texture Sharing

• Abu App Builder — Software / Authoring Tool

  All-in-one creation suite to author, package, and deploy interactive Web and Windows applications directly on top of 3D & 3DGS environments.

  Specifications: Direct SPZ / PLY / Gaussian Splat Asset Ingestion | Full WebGPU / WebGL Shader Pipeline | One-Click Windows .exe Standalone Packaging | Multi-User WebSocket Cloud Collaboration

  Protocols: Universal Scene Description (USD / OpenUSD) | glTF 2.0 / GLB with PBR Material Extensions | WebAudio Spatial 3D Sound | REST & GraphQL Content Sync

  Engines: PlayCanvas Engine 2.0 | Three.js / WebGPU Compute | Next.js App Router Web Apps | Electron / Chromium Native Shell

• Abu Cam — Software / Blender Add-on & 3DGS Pipeline

  All-in-one production camera rig (Orbital & Volumetric multi-camera dome), procedural Voronoi environment sky sphere, GPU multi-pass reference renderer (RGB, 16-bit Depth, Normal, Wireframe), and automated 3DGS / COLMAP / RealityScan / NeRF dataset exporter for Blender 4.2+ & 5.2 LTS.

  Specifications: Dual Capture Systems: 360° Orbital Turntable Rig & Volumetric Multi-Camera Spherical Dome (Fibonacci Spiral, Latitude Rings, Stratified) | Pro Environment Sky Sphere (AbuCam_EnvSkySphere) with 6 Shader Modes (Default 3D Voronoi Feature-Matching Preset, HDRI Panorama, Nishita Sky, Cyclorama Gradient) | Simultaneous 4-Pass GPU Reference Engine: 8-bit RGB Beauty, 16-bit Linear Depth (65,535 levels), Camera-Space Surface Normals & Anti-Aliased Wireframe | Independent Sensor W x H (mm) Optics, Focal Length, Live Resolution Sync & Non-Blocking Telemetry Render Monitor

  Protocols: COLMAP PINHOLE Sparse Reconstruction (cameras.txt, images.txt, binary points3D.ply sampled from mesh surface) | RealityCapture / RealityScan Exact Pose Schema (cameras.csv & camera_poses.csv with 35mm Equivalent & Distortion) | NeRF & 3D Gaussian Splatting Camera Intrinsics & 4x4 Transforms (transforms.json) | Automated FFmpeg H.264 MP4 Multi-Pass Video Encoding & glTF 2.0 / GLB Scene Preview Export

  Engines: Blender 4.2 LTS, 5.1 & 5.2 LTS (EEVEE Next Real-Time GPU & Cycles NVIDIA OptiX RTX) | 3DGS & Radiance Field Trainers (Postshot, LichtFeld Studio, Nerfstudio) | Photogrammetry & SfM Suites (Epic Games RealityScan / RealityCapture, COLMAP) | AI Video & ControlNet Generation Pipelines (Wan, ComfyUI, Multi-Pass Conditioning)

• Abu Align & Train — Software / Neural Photogrammetry & 3DGS Studio

  Unified, free, and open-source desktop studio that brings AI-driven camera alignment (VGGT-Omega, NVIDIA VGG-T3, Niantic ACE0, cuSFM, OpenSfM), 360° spherical photogrammetry, dataset preparation, and real-time 3D Gaussian Splatting (3DGS) training into a single one-click application.

  Specifications: Multi-Engine Camera Alignment: VGGT-Omega & NVIDIA VGG-T3 (Feed-Forward 3D Transformers), Niantic Labs ACE0 (ACEZero), NVIDIA Isaac cuSFM (PyCuSFM) & COLMAP | 360° Spherical Photogrammetry: Mapillary OpenSfM & 360DVO with 360° Panoramic Virtual Rig Splitter for Insta360, Ricoh Theta & GoPro Max | Direct 3D Gaussian Splatting (3DGS) Training: LichtFeld Studio (Real-Time Vulkan + GUT Lens Distortion), Nerfstudio (Splatfacto), NVIDIA 3DGRUT & PFGS360 | Complete Pre-Processing & Manual Control Suite: 4K/8K/360° Video Frame Extractor & Sharpness Picker, SAM 2 AI Sky/Object Masker & Interactive Manual Control Points (Abu M54)

  Protocols: Standardized 3dgs_dataset/ One-Click Export Pipeline | COLMAP & Nerfstudio (sparse/0/, transforms.json & dense .ply point clouds) | RealityScan / RealityCapture Exact Camera Rigs (.csv & .xml XMP camera rigs) | Native .ply & .splat 3D Gaussian Splatting Model Generation

  Engines: Neural & Metric Alignment: VGGT-Omega, NVIDIA VGG-T3 (VGG-TTT), Niantic ACE0 (ACEZero), NVIDIA cuSFM, COLMAP, Mapillary OpenSfM & 360DVO | 3DGS Training: LichtFeld Studio (Vulkan), Nerfstudio (Splatfacto), NVIDIA 3DGRUT & PFGS360 | Universal Studio Handoff: Jawset Postshot, Epic Games RealityScan / RealityCapture & Nerfstudio | Cross-Platform Single-File Installers: Windows 10/11 x64 (.exe), Linux x86_64 (.AppImage) & macOS Universal (.dmg)

Visual & Spatial Engineering Solutions

• Applied Vision & Spatial AI

  Neural Radiance & 3D Gaussian Splatting Training, AI-Assisted Camera Calibration & Alignment, Monocular Depth Estimation & Mesh Generation, Custom Vision Pipeline Integration, Synthetic Data & Neural Pre-visualization — Intelligent computer vision pipelines tailored for production workflows.

• XR & Virtual Production

  Real-Time Unreal Engine 5 Stage Integration, In-Camera VFX (ICVFX) System Architecture, Multi-Camera Genlock & PTP Synchronization, Optical Tracking Calibration & System Tuning, Stage Handover & Crew Operational Training — Turnkey stage setup with full operational handover.

• Spatial Capture & 3DGS

  High-Density LiDAR Scanning & Photogrammetry, Multi-Angle Camera Alignment & Ground Truth Data, 3D Gaussian Splatting (3DGS) Scene Optimization, Real-World Environment & Digital Twin Capture, Mesh Simplification & LOD Pipeline Preparation — Production-grade spatial assets ready for real-time engines.

• Real-Time & Web3D Apps

  Bespoke WebGPU & 3DGS Interactive Web Viewers, Custom Tools for Unreal Engine & NVIDIA Omniverse, Cloud-Connected Real-Time Data Pipelines, Cross-Platform Authoring & Desktop Toolkits, Web3D Interactive Spatial Portals — Interactive web and native applications bridging spatial data.

R&D Lab Experiments & Development Posts (122 Published Projects)

• TextureLess Splat — Radiance Field & Volumetric 3DGS R&D

  Exploring how 3D Gaussian Splatting handles textureless subjects. So far, I’ve found that 3DGS can handle textureless subjects well, as long as the camera alignment is accurate. Pipeline: Scanning: Blender using AbuCam Add-ons Alignment: RealityScan, using camera alignment data from AbuCam. Training: Postshot, with the option to use camera poses directly from AbuCam’s CSV export. Download AbuCam Blender Add-ons from the products page. Media server powered by BTS | Behind the Stage

  Challenge: Aligning cameras in textureless subjects

  Result: 500 frames rendered using the AbuCam Blender Add-ons, along with camera alignment data exported in JSON and CSV formats.

  Technologies: 3DGS, RealityScan, PostShot, AbuCam

• AI 4DGS — Radiance Field & Volumetric 3DGS R&D

  Exploring generative AI 4DGS from a single image. The main challenge now is keeping the 240 PLY sequences accurately aligned. Pipeline: Single Image → 10-second video (240 frames): #GoogleFlow / #OmniFlash Each frame → 360° orbital video: Google Flow / OmniFlash Align one orbital sequence: #RealityScan Train all 240 orbital sequences using the first aligned sequence: #Postshot Processing powered by BTS | Behind the Stage.

  Challenge: The main challenge now is keeping the 240 PLY sequences accurately aligned.

  Result: A 4DGS prototype

  Technologies: 3DGS, RealityScan, PostShot

• Makeup 3DGS — Radiance Field & Volumetric 3DGS R&D

  Exploring GenAI-driven 3D Gaussian Splatting (3DGS) digital human heads for the beauty and makeup industry. Transforming a single 2D image into an interactive web experience. End-to-End Pipeline: Concept Image: Nano Banana / Google Flow 360° Motion Reference: Blender Camera Orbit Synthesis: OmniFlash / Google Flow Structure from Motion (SfM): RealityScan (Camera Poses & Point Cloud) 3DGS Training: LichtFeld Studio Cinematic Rendering: Blender + KIRI 3DGS Add-on Interactive Web Spatial Engine: Abu Engine (PlayCanvas / Spark.js) Compute & Processing: Powered by BTS | Behind the Stage

  Challenge: Creating a clean 360° orbit while keeping her completely frozen in place with zero movement.

  Result: A 3DGS digital human head

  Technologies: 3DGS, 3DReconstruction

• Judo 3DGS — Radiance Field & Volumetric 3DGS R&D

  A Judo 3DGS scene accelerated by AI. Pipeline: Reference Image: Nano Banana 2 Lite 3D Mesh & Pose: SAM 3D Camera Reference: Blender 360° Orbit Video: OmniFlash Camera Alignment & Point Cloud: RealityScan Training: LightField Web Viewer: Abu Engine (PlayCanvas & Sparkjs Splat Renderer) Processing powered by BTS.

  Challenge: Creating a clean 360° orbit while keeping both judokas completely frozen in place with zero movement.

  Result: A 3D Gaussian Splatting Scene out of a Judo Scene

  Technologies: 3DGS, RealityScan, PostShot

• Telegram Photo to 3DGS — Radiance Field & Volumetric 3DGS R&D

  My friend sent me a photo on Telegram, so I turned it into a #3DGS. Workflow: 360° Orbital Video: #GoogleFlow / #OmniFlash Frame Alignment: #RealityScan Training: #PostShot Processing powered by BTS | Behind the Stage .

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: 3DGS, GoogleFlow, OmniFlash, RealityScan, PostShot

• AI-Assisted 3DGS Synthesis — Radiance Field & Volumetric 3DGS R&D

  Exploring #NanoBanana and #OmniFlash for AI-assisted #GaussianSplatting. This experiment starts with 4 reference images captured in the bedroom. Pipeline: - Background removal and lighting enhancement with Nano Banana - 360° orbit video generation under different lighting conditions using OmniFlash - Frame alignment and camera reconstruction in #RealityScan - Gaussian Splat training in #Postshot Processing powered by BTS | Behind the Stage. Audio: #FlowMusic

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: NanoBanana, OmniFlash, GaussianSplatting, RealityScan, Postshot

• Google Meet 3DGS Avatar — Radiance Field & Volumetric 3DGS R&D

  Creating a #3DGaussianSplatting Head from a Google Meet Screenshot. Workflow: Initial image: Google Meet screenshot of me and Bartosz Jankowski. First frame generation: #NanoBanana2Lite (#GoogleFlow) Orbit video: #OmniFlash (GoogleFlow) Image alignment: #RealityScan #3DGS training: #Postshot Processing powered by BTS | Behind the Stage.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: 3DGaussianSplatting, NanoBanana2Lite, GoogleFlow, OmniFlash, RealityScan

• Voice Cloning with GoogleFlow — Real-Time Tracking & VP Pipeline

  Exploring and developing #voicecloning with #GoogleFlow. This video was created using the #OmniFlash model together with a custom voice cloning tool and a few additional mixing enhancements. I used a single random WhatsApp voice message from Bartosz Jankowski as the voice reference, along with one screenshot of him from a meeting. In the prompt, I asked the model to have him talk about #manifold math, get out of a car, and walk into a faculty classroom. You can try this custom Flow FX tool using this link: https://lnkd.in/db9aPmHp Processing powered by BTS | Behind the Stage.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: voicecloning, GoogleFlow, OmniFlash, manifold

• GoogleFlow VFX Compositing — Real-Time Tracking & VP Pipeline

  Developing AI-accelerated video #compositing. Shot on an iPhone 16 Pro Max in Log and composited into a gas station reference scene using #GoogleFlow #OmniFlash for AI-powered masking, relighting, and scene composition. Processing powered by BTS | Behind the Stage.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: compositing, GoogleFlow, OmniFlash

• Sony Ocellus Tracking Setup — Real-Time Tracking & VP Pipeline

  Testing and developing the Sony #Ocellus Tracking System at Garage Studio and BTS | Behind the Stage. Mainly focusing on stress-testing the tracking stability under changing lighting conditions. There are no specific reflective tracking markers involved. The system performs tracking based on natural features within the environment. The lens distortion calibration was not applied yet, which led to some drift in alignment. Watch in 4K on YouTube: https://lnkd.in/gFesjrU4

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Ocellus

• Studio Green Screen Calibration — Real-Time Tracking & VP Pipeline

  Exploring the possibilities of #GreenScreen #VirtualProduction on a Cliff in #RealTime. The camera movement is pre-programmed, while the #3D background and set extension garbage matte are generated in #UnrealEngine 5.7. Lighting is controlled via #DMX inside Unreal Engine, and the #Compositing and #Colorkeying are done using Blackmagic Ultimatte 4K. Camera #tracking is sent over #FreeD using the G-ka mechanical tracking system. Developed and filmed at Garage Studio and BTS | Behind the Stage, #Dubai. Watch in 4K on YouTube: https://lnkd.in/dR8zSy5P

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GreenScreen, VirtualProduction, RealTime, UnrealEngine, DMX

• Sony FR7 AI Auto-Framing — Real-Time Tracking & VP Pipeline

  Testing #Sony FR7 #AI Auto Framing – Just by Upgrading to the Latest Firmware V4.00 This technology, known as AI PTZ Auto Framing, uses deep learning to detect human faces, allowing smooth and natural tracking. I should add that with this firmware, the FR7 can send #PTZ tracking data and Lens Metadata to #UnrealEngine via the #FreeD / #OpenTrackIO protocol. In this test, three Sony #FR7 cameras are controlled over IP using their default HTML web interface. I selected my face as the tracking target in all three cameras and then tried to challenge the system by moving quickly, rotating, and occasionally hiding my face to see how precise the auto-framing feature is. The cameras are set up with different lenses: One with an 85mm lens The other two with 16mm and 24mm lenses Developed and filmed in the Garage Studio Podcast Room and BTS | Behind the Stage, #Dubai.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Sony, PTZ, UnrealEngine, FreeD, OpenTrackIO

• Blackmagic & G-Ka LiveLink — Real-Time Tracking & VP Pipeline

  Here is the breakdown of a #GreenScreen #VirtualProduction setup in #UnrealEngine 5.7, including #RealTime #Set_Extension. The setup uses Blackmagic equipment together with the G-ka robotic arm mechanical camera tracking system using #FreeD data. Filmed and developed at Garage Studio and BTS | Behind the Stage in #Dubai.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GreenScreen, VirtualProduction, UnrealEngine, RealTime, Set_Extension

• WorldLabs Marble in UE5 — Radiance Field & Volumetric 3DGS R&D

  Exploring WorldLabs #Marble for #GaussianSplatting within a #virtualproduction workflow using #UnrealEngine 5.7 and the camera motion control tracking system. First, the reference image is enhanced using Google #NanoBanana Pro. From there, WorldLabs Marble handles the automated #3DGS generation pipeline based on the processed reference. The Marble system appears to involve AI-generated 360 imagery, depth mapping, constrained point clouds, camera projections, and Gaussian Splatting training. Tested and developed at Garage Studio & BTS | Behind the Stage Dubai. G-ka Robotic Arms. #GreenScreen Check the 3DGS background: https://lnkd.in/dD9W-4bG Check the Video in 4K: https://lnkd.in/dgha8shD

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Marble, GaussianSplatting, virtualproduction, UnrealEngine, NanoBanana

• Real-Time G-Ka Tracking — Real-Time Tracking & VP Pipeline

  Here’s the development of #RealTime mechanical tracking in #VirtualProduction, using the G-Ka robotic arm and rail system, with FreeD tracking data streamed into #UnrealEngine and #nDisplay. Garage Studio, Dubai, and BTS | Behind the Stage. In Collaboration with G-ka and Future Art Broadcast Trading L.L.C The 3D Scene designed by Sattar K.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: RealTime, VirtualProduction, UnrealEngine, nDisplay

• 3DGS in Unreal Engine 5.7 — Radiance Field & Volumetric 3DGS R&D

  Exploring 3D #GaussianSplatting in #GreenScreen #VirtualProduction using #UnrealEngine 5.7 and the #Postshot plugin, with Real-Time mechanical tracking via the G-ka Rolly robot arm. At Garage Studio, BTS | Behind the Stage and Future Art Broadcast Trading L.L.C . Scan details: • ~2,000 images captured on Nikon Z9 (50MP) • Aligned in #RealityScan • Trained in #Postshot View the interactive #3DGS here: https://lnkd.in/d99Uu4tY

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, GreenScreen, VirtualProduction, UnrealEngine, Postshot

• NanoBanana Pro & Veo3 — Real-Time Tracking & VP Pipeline

  #AI Multi-Angle Scene Continuity Study, Using #NanoBananaPro and #Veo3 Exploring AI-driven continuity across multi-angle shots within a single environment. Using reference photos of myself and various store perspectives. Processing backed by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: NanoBananaPro, Veo3, nanobananapro, veo3

• Gemini 3 + NanoBanana Pro — Radiance Field & Volumetric 3DGS R&D

  Exploring #NanoBananaPro and #Veo3.1 using a single image I captured tonight with the iPhone 16. The entire video was created on mobile in just minutes — from taking the photo, to generating with #Gemini3, and finally editing and exporting this breakdown using the Premiere mobile app. The game is getting interesting ;)

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: NanoBananaPro, Veo3, Gemini3, nanobananapro, veo3

• Veo3 AI VFX Composition — Real-Time Tracking & VP Pipeline

  Exploring #Veo3, #NanoBanana, and #Photoshop in #AI-accelerated #VFX composition. Processing backed by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Veo3, NanoBanana, Photoshop, VFX, veo3

• PointCloud & 3DGS Fusion — Radiance Field & Volumetric 3DGS R&D

  Here’s the breakdown of using Generative AI in VFX composition, along with a 3D #GaussianSplatting point cloud comparison against the final renders. Processing backed by BTS | Behind the Stage #Veo3 #NanoBanana #Photoshop #PostShot #RealityScan

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Veo3, NanoBanana, Photoshop, PostShot

• Alif Pavilion GenAI Video — Radiance Field & Volumetric 3DGS R&D

  Here’s a use case of #GenAI + 3D #GaussianSplatting for video creation, showcasing the Alif Pavilion at Expo #Dubai. Splat Positioning: #RealityScan Splat Training: #PostShot, #Marble World Labs AI Video: #Veo3 AI Image Composition: #NanoBanana, #Photoshop Voice Over: Artlist.io Processing backed by: BTS | Behind the Stage #AlifChallenge

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GenAI, GaussianSplatting, Dubai, RealityScan, PostShot

• Address Al Marj Visualizer — Real-Time Tracking & VP Pipeline

  Here’s the breakdown using #NanoBanana, #Veo3, and #Photoshop to create an #AI video from rendered images of an off-plan property. The visuals are based on the rendered photos of Address Al Marjan Island by Emaar in Ras Al Khaimah, UAE. 🔗 Reference images: https://lnkd.in/ddPY8vvr Processing powered by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: NanoBanana, Veo3, Photoshop

• Photoshop NanoBanana AI — Real-Time Tracking & VP Pipeline

  Showcasing Off-Plan Properties with #AI ✨ This video was generated from rendered images of ADDRESS AL MARJAN ISLAND by Emaar in Ras Al Khaimah, using #Photoshop, #Nanobanana and #Veo3. Powered by BTS | Behind the Stage media server.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Photoshop, Nanobanana, Veo3, photoshop, nanobanana

• Broadcast AR Layers in UE5 — Real-Time Tracking & VP Pipeline

  In-progress look at developing #AR Layers for #broadcasts: using two iOS devices — one as the camera and the other for tracking over Wi-Fi — with Unreal Engine 5.6’s native #VirtualProduction features. Once refined, this workflow will be put into action at Ascend Esports for game broadcasts — special shoutout to the team and this collaboration. Processing powered by BTS | Behind the Stage.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: broadcasts, VirtualProduction

• AI Characters in Real Estate — Radiance Field & Volumetric 3DGS R&D

  Exploring the blend of 3D #GaussianSplatting #NanoBanana #Veo3 in real estate promotions. All four characters in this video are fully #AI-generated—from video and audio to their appearance and dialogue—assisted by #Gemini . 📸 Photo scanned with iPhone 16 using 2,553 images 📍 Positioned by RealityScan 🖥️ Trained on PostShot ⚙️ Processing backed by BTS | Behind the Stage ✨ Check out the splats gallery: https://lnkd.in/dCRrVrhe DAMAC Properties #Dubai

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, NanoBanana, Veo3, Gemini, Dubai

• NanoBanana Image Generation — Real-Time Tracking & VP Pipeline

  Exploring Google’s #NanoBanana and #Veo3 📸 Snapped a quick photo of the bag and chair with my phone. 🐱 Added the cat into the scene using NanoBanana. 🎬 Generated both the video and audio in Veo3, starting from the composited image as the first frame.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: NanoBanana, Veo3, nanobanana, veo3

• 3DGS vs 360°° Photography — Radiance Field & Volumetric 3DGS R&D

  When I show 3D #GaussianSplatting samples—even in #VR — people often confuse them with #360° photos. So I made this comparison. Hope it helps explain the medium better! Feel free to share it with anyone who’s had the same confusion. 📸 Scan: 834 photos with iPhone camera 📍 Positioned in RealityScan ⚙️ Trained on #PostShot 💻 Processing backed by BTS | Behind the Stage 👉 Explore this 3D splat here: https://lnkd.in/dCRrVrhe Using #SuperSplat Viewer

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, 360, PostShot, SuperSplat

• Varjo Teleport 3DGS Viewer — Radiance Field & Volumetric 3DGS R&D

  Reviewing smart unit setups inside the bus using 3D #GaussianSplatting on Teleport by Varjo Web Viewer. Scanned with 1,876 images on iPhone Positioned in RealityScan Trained with #Postshot Processing powered by BTS | Behind the Stage Explore this splat in 3D: https://lnkd.in/dCRrVrhe

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Postshot

• Wan22 AI Scene Harmonize — Real-Time Tracking & VP Pipeline

  Exploring #AI tools to bring my dreams to life. I first saw this scene in a dream two years ago—now recreated through technology. Brought to life using #Gemini, #PhotoshopHarmonize, #ComfyUI, #WAN22 #AI processing powered by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Gemini, PhotoshopHarmonize, ComfyUI, WAN22, gemini

• Dubai PostShot 3DGS Splat — Radiance Field & Volumetric 3DGS R&D

  Exploring Google #VEO3 with 3D #GaussianSplatting at the MOHAMMED BIN RASHID LIBRARY, #Dubai 📸 Photoscanned with 2,115 images on iPhone 14 Positioned in RealityScan Trained on #PostShot Composite light matching done in #Photoshop using the #Harmonize AI feature, and the video composition done in #AfterEffects 🔗 View this splat in 3D (with #SuperSplat viewer): https://lnkd.in/dCRrVrhe 🎥 Watch in HQ on YouTube: https://lnkd.in/d7shURBy Processing backed by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VEO3, GaussianSplatting, Dubai, PostShot, Photoshop

• Veo3 Generative Motion — Real-Time Tracking & VP Pipeline

  Turning My Photo + Speech Text into a Video with #VEO3 Supported by BTS | Behind the Stage Watch on YouTube: https://lnkd.in/eQ7RMkSK

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VEO3, veo3

• Google Veo3 Scene Harmonize — Radiance Field & Volumetric 3DGS R&D

  Exploring the new #AI features in #Photoshop (#Harmonize) and #Google #VEO3 for creating stunning images and videos. The bus image was captured at Al Khail Bus Station in #Dubai, and I borrowed my daughter’s doll for the shoot. It was such an interesting experience—kids around were amazed and excited to see their toys come to life through animation, turning their imaginations into reality! 🎵 Music by Benjamin Tissot Supported by BTS | Behind the Stage 📺 Watch the video on YouTube: https://lnkd.in/daH9GVv7

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Photoshop, Harmonize, Google, VEO3, Dubai

• iPad 3DGS in Unreal Engine — Radiance Field & Volumetric 3DGS R&D

  Exploring 3D #GaussianSplatting in a remote session using an #iPad and #UnrealEngine 5.6 #LiDAR Scanning using XGRIDS L2PRO Aligned & trained by #LixelCyberColor Location: Garage Studio Scanner provided by UNIFY TEC Processed on BTS | Behind the Stage media servers 🔗 View the 3D environment: ahmad-emranian.netlify.app 🎥 Watch in 4K on YouTube: https://lnkd.in/d3G4EA-a

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, iPad, UnrealEngine, LiDAR, LixelCyberColor

• Gemini + Photoshop AI Pipeline — Radiance Field & Volumetric 3DGS R&D

  Exploring a large 3D scene with #GaussianSplatting using the XGRIDS L2PRO scanner. The studio was scanned in four separate sections and then combined into a single splat for seamless visualization. Alignment & Training: #LixelCyberColor Camera Animation: #UnrealEngine 5.6 Scanner Provided by: UNIFY TEC Processing by: BTS | Behind the Stage Location: Garage Studio 🔗 View the 3D environment: ahmad-emranian.netlify.app 🎥 Watch in 4K on YouTube: https://lnkd.in/dPUgHc5c Feel free to reach out for projects or inquiries 🙌

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, LixelCyberColor, UnrealEngine, gaussiansplatting, lixelcybercolor

• Lixel CyberColor Handheld Scan — Radiance Field & Volumetric 3DGS R&D

  Exploring XGRIDS-Scanned Environments Enhanced with Photoscanned Subjects in 3D #GaussianSplatting. The apartment was scanned using the XGRIDS L2 Pro and processed with #LixelCyberColor. Additionally, the frames, flower table, and PC desk were photoscanned with an #iPhone (1,203 images), aligned in RealityScan, and then trained and composited in #PostShot. XGRIDS Scanner provided by UNIFY TEC Processing by BTS | Behind the Stage 🔗 Experience this mixed splat on desktop via #SuperSplat and Teleport by Varjo Teleport: https://lnkd.in/dCRrVrhe

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, LixelCyberColor, iPhone, PostShot, SuperSplat

• AI Metahuman Real Estate Host — Radiance Field & Volumetric 3DGS R&D

  Exploring #LiDAR in 3D #GaussianSplatting with the XGRIDS L2 Pro Scanner. Laser scanner: Provided by UNIFY TEC Processing: Done by BTS | Behind the Stage Location: PHYSIOLAB CENTRE LLC (Thanks to Daniel D.) 🔗 View the 3D environment: ahmad-emranian.netlify.app 🎥 Watch in 4K on YouTube: https://lnkd.in/dRHFQn67 And Thanks to Dano Joseph Cherison for the support!

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: LiDAR, GaussianSplatting, lidar, gaussiansplatting

• Veo3 Dubai Architectural Splat — Radiance Field & Volumetric 3DGS R&D

  Exploring 3D #GaussianSplatting for #RealEstate visualization. View the interactive demo on desktop via Varjo Teleport: https://lnkd.in/dNTb8rcq Click “Start Tour” and use the portals to explore the spaces. 📸 Captured with 5,646 images (iPhone 16) 🧠 Aligned in RealityScan ⚙️ Trained with #Postshot #AI Processing BTS | Behind the Stage Watch in 4K on YouTube: https://lnkd.in/dQV9eJ7P

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, RealEstate, Postshot, gaussiansplatting, realestate

• Veo3 Volumetric Motion Test — Radiance Field & Volumetric 3DGS R&D

  Exploring the use of 3D #GaussianSplatting in game development with #UnrealEngine 5.6 I had my son, Ali, run the UE Shooter sample project inside this industrial warehouse I scanned — adding a few simple invisible meshes for collision on walls and the ground. 📸 Scanned using 1369 images with the DJI Osmo 1 🧠 Aligned and trained with #PostShot #AIProcessing by BTS | Behind the Stage Watch on YouTube 4K: https://lnkd.in/dFJj8axf

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, UnrealEngine, PostShot, AIProcessing, gaussiansplatting

• LiDAR & 3DGS Integration — Radiance Field & Volumetric 3DGS R&D

  Eliminating the shade on the building in 3D #GaussianSplatting by combining two photogrammetry scans taken at different times of day — resulting in a sharper, more detailed reconstruction. Morning scan: 1,127 images Afternoon scan: 938 images All 2,065 images captured with iPhone 16 Aligned with RealityScan Trained using #PostShot Check it out in 3D on #SuperSplat: https://lnkd.in/dXSqa-8n Watch in 4K on YouTube: https://lnkd.in/ds-Sfd3s #AI Processing BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, PostShot, SuperSplat

• XGRIDS Apartment Scan — Radiance Field & Volumetric 3DGS R&D

  Exploring camera and interaction settings, along with graphic compositing on web and 3D #GaussianSplatting — using Spline's web editor and player. This recorder was scanned with 892 images using iPhone 16,  Aligned in #RealityScan,  And trained with #PostShot. #AI Processing BTS | Behind the Stage  Check it out here: 🔗 #Spline: https://lnkd.in/dd2TK6sJ 🔗 #SuperSplat: https://lnkd.in/dbzwhZkv 🔗 Youtube 4k: https://lnkd.in/dGvx7G4u

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, RealityScan, PostShot, Spline, SuperSplat

• Real Estate RealityScan 3DGS — Radiance Field & Volumetric 3DGS R&D

  Exploring camera animation with 3D #GaussianSplatting in video creation. This entire video is a 3D-rendered animation in #UnrealEngine — no footage was used. Captured with 3,525 photos using iPhone 16 Aligned in #RealityCapture Trained using #PostShot Studio host: Mark Al Kadamani #AI Processing BTS | Behind the Stage 🎥 Watch in higher quality: https://lnkd.in/d_NfiJuf

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, UnrealEngine, RealityCapture, PostShot, gaussiansplatting

• PostShot Radiance Optimizer — Radiance Field & Volumetric 3DGS R&D

  Exploring 3D #GaussianSplatting in #VirtualProduction with #GreenScreen and #Realtime #VFX using cinematic gear in 4K and #UnrealEngine. Scanned with 357 images using #iPhone16 Aligned in #RealityCapture Trained in #PostShot Grateful to: StreamPort Media – VIVE MARS tracking Mark Al Kadamani – Studio host BTS | Behind the Stage – On-set #VP servers & #AI processing 🔗 Watch in higher quality: https://lnkd.in/dFV6HU_7

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VirtualProduction, GreenScreen, Realtime, VFX

• Multi-Scan Fusion 3D Model — Radiance Field & Volumetric 3DGS R&D

  #GaussianSplatting for Product Promotion Video. Captured with 243 photos on iPhone 16 Aligned and trained with #PostShot Rendered in #UnrealEngine 5.6 AI processing by BTS | Behind the Stage #3DGS

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, PostShot, UnrealEngine, 3DGS, gaussiansplatting

• PostShot UE5 VP Pipeline — Radiance Field & Volumetric 3DGS R&D

  How can this shot be achieved in the real world? Exploring 3D #GaussianSplatting in video effects. Captured with 166 images using an iPhone 16 at Burj Khalifa, Dubai. Aligned in #RealityCapture. Trained with #PostShot. #AI processing by BTS | Behind the Stage Check out my splats: https://lnkd.in/dCiu2pG7 Purchase my splats: https://lnkd.in/dxrsiKti

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, RealityCapture, PostShot, gaussiansplatting, realitycapture

• RealityCapture to Splat — Radiance Field & Volumetric 3DGS R&D

  Exploring 3D #GaussianSplatting in #VFX. The Bus #3DGS was created from 151 images captured using iPhone 16. Positioned in #RealityCapture Trained with #PostShot Composited in #AfterEffects AI Processing - BTS | Behind the Stage 🛒Purchase and download this and my other Splats: https://lnkd.in/dxrsiKti

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VFX, 3DGS, RealityCapture, PostShot

• Virtual Production Splatting — Radiance Field & Volumetric 3DGS R&D

  Exploring a Full 360° Product View Using 3D #GaussianSplatting Check it out on supersplat here: https://lnkd.in/dPJir7qV Based on 951 images captured with iPhone 14 #AI processing: BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, gaussiansplatting

• PostShot Unreal Engine Sync — Radiance Field & Volumetric 3DGS R&D

  3D #GaussianSplatting for #RealEstate This #3DGS captures the SOBHA Realty maquette display at Mall of the Emirates, Dubai—scanned with 2501 images using a Nothing Phone 1. 🔹 Camera Positioning: #RealityCapture 🔹 Splat Training: #postshot 🔹 Composition: #UnrealEngine Check this Demo in 3D on #SuperSplat (PlayCanvas): https://lnkd.in/dyNcZngB Check out more splats: saemranian.vercel.app #AI Processing by BTS | Behind the Stage 💡 Open to remote 3DGS guidance & processing.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, RealEstate, 3DGS, RealityCapture, postshot

• UE5 Photogrammetry Alignment — Radiance Field & Volumetric 3DGS R&D

  3D #GaussianSplatting in #Realtime Composition | #VFX #VirtualProduction #GreenScreen. The 3DGS gallery background is running in 6K Resolution at 25 FPS with a single RTX 4090, Using Unreal Engine and the #PostShot plugin, with VIVE MARS Tracking (tracking alignment was improved later). Developed in Garage Studio and BTS | Behind the Stage. The BG is Mustapha Azab Gallery At #Xposure 2025, #Sharjah.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Realtime, VFX, VirtualProduction, GreenScreen

• Generative AI 3DGS Synthesis — Radiance Field & Volumetric 3DGS R&D

  Developing 3D #GaussianSplatting in the #GreenScreen #VirtualProduction setup at Garage Studio and BTS | Behind the Stage. #Realtime #Tracking: VIVE MARS Realtime #VFX Composition: #EnrealEngine 5.5 + #BlackMagic #Ultimatte Background Scanned: 603 images with #Nikon Z6 Camera Positioning and Point Cloud: #RealityCapture

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, GreenScreen, VirtualProduction, Realtime, Tracking

• After Effects 3DGS VFX — Radiance Field & Volumetric 3DGS R&D

  Sony ALPHA FESTIVAL in 3D #GaussianSplatting This Video is from the Sony MEA ALPHA FESTIVAL, held on January 18th at Garage Studio in Dubai. Captured with 2552 RAW images in 8K at 30 fps (Burst Mode). Positioned in #RealityCapture. Trained and Composited on #PostShot #AI Processing by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, RealityCapture, PostShot, gaussiansplatting, realitycapture

• 6K 3DGS Gallery at 25 FPS — Radiance Field & Volumetric 3DGS R&D

  Compositing 4D and 3D #GaussianSplatting in Video #VFX. 🛹 The Skater #VolumetricVideo was captured and processed by GeniusXR and UVRSE in Canada using 74 Sony RX0 II cameras in 4K at 30 FPS. 🏙️ The Historic Alley in Dubai was captured with 1,033 images using an iPhone 14 Pro Max, positioned in #RealityCapture, trained on #PostShot, and processed on BTS | Behind the Stage media servers. 🎬 Final compositing done in #PostShot.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VFX, VolumetricVideo, RealityCapture, PostShot

• Real Estate 3DGS Showcase — Radiance Field & Volumetric 3DGS R&D

  Pre-visualizing Building Facade Lighting Using 3D #GaussianSplatting and #UnrealEngine 5.5. This experiment is ALDAR Building in Abu Dhabi. Captured with iPhone 16 Pro Max (539 images). Aligned in #RealityCapture | Trained with #PostShot Check out more splats: saemranian.vercel.app #AI Processing by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, UnrealEngine, RealityCapture, PostShot, gaussiansplatting

• RealityCapture 3DGS Splat — Radiance Field & Volumetric 3DGS R&D

  Exploring 3D #GaussianSplatting as #AR layers in #VirtualProduction with #UnrealEngine 5.5 at Garage Studio and BTS | Behind the Stage in Dubai. Using VIVE MARS Tracking system in #Realtime. Head and Face Scanning: - Captured with 428 images using a #Nikon Z9 in burst mode at 30 fps. - In 8K RAW Resolution. - I asked the Model to freeze, and completed the scan in less than 1 minute. Processing Workflow: - Aligned and positioned in #RealityCapture. - Trained with #PostShot. Check out this splat: https://lnkd.in/dJDeUwrb Explore more splats: https://lnkd.in/dCxGbq3s

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VirtualProduction, UnrealEngine, Realtime, Nikon

• Volumetric Night Splatting — Radiance Field & Volumetric 3DGS R&D

  3D #GaussianSplatting at Night. #PhotoScanning with 414 images using an iPhone 16 Pro Max. Positioned in #RealityCapture. Trained on #PostShot. Check out my splats: https://lnkd.in/dCxGbq3s #AI Processing by BTS | Behind the Stage.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, PhotoScanning, RealityCapture, PostShot, gaussiansplatting

• RealityScan UE5 LiveLink — Radiance Field & Volumetric 3DGS R&D

  Translucency in 3D #GaussianSplatting A #MixedReality experience featuring Gracia AI Captured using 370 images on an iPhone 15 Pro Max Positioned in #RealityCapture Trained with #PostShot Thanks to Till Louesse for #MetaQuest3 #AI Processing by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, MixedReality, RealityCapture, PostShot, MetaQuest3

• Green Screen VP Splatting — Radiance Field & Volumetric 3DGS R&D

  3D #GaussianSplatting of #Sharjah Mosque. Photoscanned with 6,680 images using an iPhone 16 Pro Max, positioned in #RealityCapture, and trained with #Postshot. The car drive test is showcased in #UnrealEngine 5.5, #MixedReality powered by Gracia AI and #MetaQuest2. Check out this splat here: Arrival.Space: https://lnkd.in/d8Qunh5P Gracia AI: https://lnkd.in/d_cPXsg3 My Website: https://lnkd.in/dCxGbq3s Wishing you all a Happy New Year! AI Processing by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Sharjah, RealityCapture, Postshot, UnrealEngine

• McLaren 720S MR on Quest 3 — Radiance Field & Volumetric 3DGS R&D

  #McLaren 720s 3D #GaussianSplatting in #MixedReality on #Quest3 using Gracia AI. Captured by 254 images using nothing phone 1 camera. Tracked in #RealityCapture and Trained on #PostShot. Special Thanks to Till Louesse for his #MetaQuest3. Check out my splats here: https://lnkd.in/dCxGbq3s AI Processing by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: McLaren, GaussianSplatting, MixedReality, Quest3, RealityCapture

• AR Broadcast Spatial Tracking — Radiance Field & Volumetric 3DGS R&D

  Honda Bike 3D #GaussianSplatting in #MixedReality. Captured using an #iPhone 14 with 2,153 images. Positioned and trained with #PostShot. The #MR experience is powered by Gracia AI Win software and displayed on #MetaQuest2. Try This Honda Splat in MR here: https://lnkd.in/dwaEk5Jn Check out my splats here: https://lnkd.in/dCxGbq3s DM me for processing your datasets. AI Processing by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, MixedReality, iPhone, PostShot, MetaQuest2

• iPhone 414-Photo Scanning — Radiance Field & Volumetric 3DGS R&D

  Dubai Museum of the Future 3D #GaussianSplatting in #MixedReality using Gracia AI and #Metaquest2. Captured with 1,100 images using the Nothing Phone 1 camera. Processed with #RealityCapture for tracking and trained in #PostShot. Explore this #3DGS: For #MR & #VR: https://lnkd.in/dPNaTzW8 For web browser: https://lnkd.in/dCxGbq3s AI processing by BTS | Behind the Stage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, MixedReality, Metaquest2, RealityCapture, PostShot

• Mixed Reality Quest 3 Setup — Radiance Field & Volumetric 3DGS R&D

  #GaussianSplatting in #VirtualProduction. Each 3D scene in this video was created using 103, 60, and 121 photos, all captured with the Nothing Phone 1 camera. Using #UnrealEngine 5.4, #RealityCapture, #PostShot and its UE plugin, VIVE MARS Tracking, and a Sony Burano, the project runs in 4K resolution at 22–25 FPS on a single RTX 4090, pushing the limits in an experimental setup. This Development is taking place in Garage Studio and BTS | Behind the Stage Virtual Production Studio. Check out my splats here: https://lnkd.in/dCxGbq3s

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VirtualProduction, UnrealEngine, RealityCapture, PostShot

• Sharjah Heritage 3DGS — Radiance Field & Volumetric 3DGS R&D

  Finally, I’m exploring my #Gaussiansplatting model in Unreal Engine 5.4, in high quality with the #Postshot UE plugin. This is the Mall of the Emirates in Dubai, which I scanned using my Nothing Phone 1 camera with 1,019 photos, positioned in #RealityCapture and trained in Postshot. I asked my son, Ali, to help me run a player within this scene using his #PS5 controller in UE 5.4. AI processing by BTS | Behind the Stage media servers. Try this splat at this link: https://lnkd.in/drfpYnNb

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Gaussiansplatting, Postshot, RealityCapture, PS5, gaussiansplatting

• Quest 3 Mixed Reality Splat — Radiance Field & Volumetric 3DGS R&D

  Experimenting with #GaussianSplatting for 3D views of car interiors. The AMG C43 #MercedesBenz at Macan Luxury Cars in Dubai was photoscanned with the Nothing Phone 1 using 700 images, while the #BMW at AGMC Dubai was photoscanned with the iPhone 14 using 105 and 163 images. Positioning done in #RealityCapture, training in #PostShot, and web browser interaction with splats using Mark Kellogg's code. #AI processing by BTS | Behind the Stage Media servers.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, MercedesBenz, BMW, RealityCapture, PostShot

• Quest 3 Hand Tracking MR — Radiance Field & Volumetric 3DGS R&D

  #GaussianSplatting in Unity. In this sample, I was trying to figure out how to get clean splats from car surface reflections, especially with darker colors. I captured many cars to understand how it works and what the restrictions are. The next step is deciding which platform to share it on or create an application from it. After testing #UnrealEngine with third-party plugins and #Unity with Aras Pranckevičius' code, I’m finally seeing my splats in a #GameEngine with nearly the same quality as I trained them. Thanks to: #RealityCapture #PostShot, #Unity, and #ArasPranckevičius #AI processing by BTS | Behind the Stage media servers.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, UnrealEngine, Unity, GameEngine, RealityCapture

• RealityCapture PostShot Splat — Radiance Field & Volumetric 3DGS R&D

  #GaussianSplatting #VFX in 4K. Here’s the new version featuring 2 little crew members, compared to the previous video I posted. All the composite layers, including the background, operators, and camera, are #Splats created from photo scans in the Garage Studio (no video footage was used in making this video). #AI processing by BTS | Behind the Stage media servers. You can also check it out on YouTube at this link: https://shorturl.at/iIe6q #RealityCapture #PostShot #AfterEffects

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VFX, Splats, RealityCapture, PostShot

• Turnkey Virtual Production — Radiance Field & Volumetric 3DGS R&D

  BMW #DigitalTwin Showroom Experience! Introducing the #GaussianSplatting model of the BMW X6 2025, created from 2,901 images captured by an iPhone 14 at the AGMC BMW showroom in Dubai. Special thanks to: Aras Pranckevičius for the #Unity splat code Mark Kellogg for the web viewer This project was aligned in #RealityCapture and trained on #PostShot. Brought to life in collaboration with Global Mobile Software LLC. #AI processing by BTS | Behind the Stage media servers.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: DigitalTwin, GaussianSplatting, Unity, RealityCapture, PostShot

• Mercedes & BMW 3DGS Demo — Radiance Field & Volumetric 3DGS R&D

  Your Construction Project at Your Fingertips! Experience 2.5M #splats in #realtime on a normal mobile browser! This #GaussianSplatting use case showcases the building development progress over 3 months. I photo-scanned this building twice over 3 months—first with 2,000 images using an iPhone, and then with 450 images using a Nothing Phone. The web splat viewer is developed by Mark Kellogg. #AI processing powered by BTS | Behind the Stage media servers #RealityCapture #Postshot

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: splats, realtime, GaussianSplatting, RealityCapture, Postshot

• Gaussian Splatting in Unity — Radiance Field & Volumetric 3DGS R&D

  6K #GaussianSplatting #DigitalTwin Experiment!   Presenting Studio 'A' at Garage Studio, where a 3x4 meter area was photo scanned using 603 images at 6K resolution.   Explore the compressed version of this Splat Model on Gracia AI with a #VR headset here: https://shorturl.at/XRGln    To view the full 6K Gaussian Splatting, watch it on my YouTube channel here: https://shorturl.at/d0yRz   #AI processing by BTS | Behind the Stage media servers.   Interested in training splats from your photoscan dataset in high resolution or starting from scratch? Let me know!   #PostShot #RealityCapture #AfterEffects

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, DigitalTwin, PostShot, RealityCapture, AfterEffects

• 4K Gaussian Splatting Video — Radiance Field & Volumetric 3DGS R&D

  Set Extension using #Gaussiansplatting! Here is The Garage Studio "A" Splat, composited with iPhone 15 footage from the Van. The photo scanning is based on 384 images. AI Processing by BTS | Behind the Stage media servers. #RealityCapture #PostShot #AfterEffects

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Gaussiansplatting, RealityCapture, PostShot, AfterEffects, gaussiansplatting

• Unity Digital Twin Splats — Radiance Field & Volumetric 3DGS R&D

  Experience the Sentient By Elysian Office in Immersive 3D with #GaussianSplatting! Special thanks to Gracia AI for their intuitive 3D Splat Viewer. The best part? You can share your splats on the web through their store, allowing anyone with the link to explore your space in VR with high quality. The first video showcases the rendered version of the #3DSplat, demonstrating the potential of scanning first and then framing your shoot! This coworking space was photoscanned with nearly 4,600 photos using an iPhone 14. AI Processing by BTS | Behind the Stage Media Servers #RealityCapture #PostShot

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, 3DSplat, RealityCapture, PostShot, gaussiansplatting

• Real-Time Splat Optimization — Radiance Field & Volumetric 3DGS R&D

  Now using #GaussianSplatting in Promotional Videos! No video was used in editing this continuous shot—all of it is based on G-Splats. The Neon Lights, Studios, Shower and Pool were photo-scanned, tracked, and trained, then composited together to create this video. As a filmmaker, #GSplats are going to be my new go-to format on my editing desk. Captured in Garage Studio and trained on BTS | Behind the Stage media servers. What is next? #RealityCapture #PostShot #AfterEffects #3DGS

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, GSplats, RealityCapture, PostShot, AfterEffects

• 6K Gaussian Splat Capture — Radiance Field & Volumetric 3DGS R&D

  The power of #GaussianSplatting in #VFX! Discover how to use G-Splats to create captivating video reels for branding and promotion on social media. All the #3D elements, including the studio, #SonyVenice2 camera, and operator, are photo scanned, tracked in #RealityCapture, trained in #PostShot, and composited in #AfterEffects. Proudly captured in the Garage Studio and trained and composited on the BTS | Behind the Stage Media Servers. Let me know what you think!

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VFX, SonyVenice2, RealityCapture, PostShot

• RealityCapture High-Res Splat — Radiance Field & Volumetric 3DGS R&D

  Another use case of #3D #GaussianSplatting in Film #VisualEffects! I photo-scanned the KIA #Pikanto using my phone camera, taking almost 1800 photos. Then, I composited it onto the footage I shot in the poolside. Let me know what you think. Trained & Rendered on BTS | Behind the Stage Media servers #RealityCapture #Postshot #AfterEffects #AI #MachineLearnign #CG #3DGS

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, VisualEffects, Pikanto, RealityCapture, Postshot

• 3DGS Volumetric Scene Graph — Real-Time Tracking & VP Pipeline

  Developing #RealTime #ObjectTracking for #AR layers in #VirtualProduction. Finally, after a year of having access to the VIVE MARS Tracking System in the Garage Studio and BTS | Behind the Stage, I decided to focus on achieving precise and stable tracking with it and it worked beautifully. This experience taught me that with a bit of patience, understanding the limitations of each device, some trial and error you can develop cost-effective solutions without breaking the bank. Virtual Production is becoming more accessible with experienced hands and efficient tools!

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: RealTime, ObjectTracking, VirtualProduction, realtime, objecttracking

• Sony Venice 2 VFX Splat — Radiance Field & Volumetric 3DGS R&D

  My favorite bus in Dubai, now in #3D #GaussianSplatting! Captured with nearly 1300 photos from my midrange mobile camera. Trained and rendered on BTS | Behind the Stage Media Servers. Let me know if there are any inquiries. #PostShot #RealityCapture #AfterEffects #3DGS #AI #AIVFX #DigitalTwin #DigitalAssets #Metaverse

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, PostShot, RealityCapture, AfterEffects, 3DGS

• RealityCapture G-Splats — Radiance Field & Volumetric 3DGS R&D

  Unleashing the Power of 3D #GaussianSplatting in #MixedReality in Sentient By Elysian Laboratory. Utilizing the potential of #3DGS for the future of #Events and #DigitalEvents. I'm impressed with how the Meta #Quest2 handles tracking and composition with #passthrough footage in real-time using #UnrealEngine. This is just the beginning! Blending the Reality with the Realistic #DigitalAssets. Trained and processed on BTS | Behind the Stage Media Servers. Special thanks to Luma AI and jawset visual computing for their Plugins and software.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, MixedReality, 3DGS, Events, DigitalEvents

• Visual Effects 3DGS Pipeline — Radiance Field & Volumetric 3DGS R&D

  3D #GaussianSplatting hits the #Metaverse! Huge thanks to Arrival.Space for making #immersive browsing a breeze. Explore my #Splat creations (and future ones!) on mobile, tablet, PC, and more! Chat with friends or me in this Metaverse space: link: https://lnkd.in/dprx-sP7 The possibilities are endless! Let me know your thoughts in the comments. Want to learn more? DM me for a course on my #GaussianSplatting pipeline for #VFX, #VR, #Metaverse & more! Open to project collabs too. Trained on BTS | Behind the Stage Media Servers. #PostShot #RealityCapture #Nvidia #AI

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Metaverse, immersive, Splat, VFX

• PostShot RealityCapture 3DGS — Radiance Field & Volumetric 3DGS R&D

  Exploring the Camel's #GaussianSplatting with just a single phone camera – no full-body #3DScan studio setup used! Here's the #3D reconstruction of FARES (the camel) from 600 photos captured on an #iPhone at the Arabian Desert Camel Riding Center (Linda Krockenberger) in #Dubai. What do you think? Let me know in the comments. Trained and rendered on the BTS | Behind the Stage's Media Servers. #PostShot #RealityCapture #AfterEffects #MachineLearning #3DBodyScanning #AI #AI3D #Nvidia

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, 3DScan, iPhone, Dubai, PostShot

• Real-Time Object Tracking AR — Radiance Field & Volumetric 3DGS R&D

  Using #GaussianSplatting, first scan your subject, then choose your frame! Cinematic look in #Dubai's sun? #VFX magic! Achieved various focal lengths & sensor sizes (Full Frame & 70mm) with cinematic movement in post. This showcases a #3DGS of a #Lamborghini #Huracan #STO trained on BTS | Behind the Stage media servers in the Garage Studio. The #Photogrammetry (#PhotoScanning) is based on 537 images captured with a midrange phone camera (#nothingphone1). Tracked with Capturing Reality, Trained with #PostShot, and composited in #AfterEffects. What possibilities does this new #AI tool hold for the future of #Media and VFX? Automobili Lamborghini S.p.A. Lamborghini Abu Dhabi & Dubai #VisualEffects #CG #RealityCapture

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Dubai, VFX, 3DGS, Lamborghini

• Mixed Reality Spatial Capture — Radiance Field & Volumetric 3DGS R&D

  Leveraging #GaussianSplatting in #VR! The first demo showcases a #DigitalTwin of the #Hafez Tomb in Shiraz, Iran, which I photogrammetered in 2019. The remaining scenes feature various industrial areas I've previously photogrammetered. Camera alignment and #AI Training were completed using #PostShot, followed by import into #UnrealEngine via the #LumaAI plugin. Having explored the #3DGS in #VirtualProduction, #VFX, and now VR, what's next? Trained on the BTS | Behind the Stage Media Servers in the Garage Studio. #MetaQuest2 #Metaverse #Web3 #VirtualReality #NvidiaAI #Photoscanning #Photogrammetry

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, DigitalTwin, Hafez, PostShot, UnrealEngine

• Metaverse Immersive Splats — Radiance Field & Volumetric 3DGS R&D

  Is The Future of MetaHumans with #GaussianSplatting and just one phone?!     Excited to share my #3DGS #DigitalTwin. Thanks to my wife for capturing a part of my face (motionless, I might add!) in the kid's room.   Huge shoutout to: Inria jawset visual computing Capturing Reality Epic Games NVIDIA AI   Trained on BTS | Behind the Stage media servers in Garage Studio   Waiting for #Dynamic3DGaussian #4DGaussianSplatting #VolumetricVideo

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, 3DGS, DigitalTwin, Dynamic3DGaussian, 4DGaussianSplatting

• 3DGS VFX Composition — Radiance Field & Volumetric 3DGS R&D

  3D #GaussianSplatting VFX #Composition Breakdown.   Here I showcase how I composited the 3D Gaussian Splattings in my recent works.   There are 3 main workflows to do that: 1. The 3D Gaussian Splatting layer on The Footage Layer 2. The Footage layer on The 3D Gaussian Splatting layer 3. The 3D Gaussian Splatting on The 3D Gaussian Splatting 4. And many more. Such as compositing with other 3D Models and Layouts, Volumetric Videos, and so on.   Trained and rendered on BTS | Behind the Stage Media servers in Garage Studio Let me know what you think in the comments. #AI #CG #AICG #AICGI #VFX #VisualEffects #AfterEffects #Jawset #Postshot #Training #Nvidia #PointCloud #Compositing #SonyVenice

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Composition, AICG, AICGI, VFX

• Dubai Real Estate 3DGS — Radiance Field & Volumetric 3DGS R&D

  And another #VFX sample powered by #AI!! Let's Shoot with the #3DGaussianSplatting Version of #SonyVenice2 in the Garage Studio and BTS | Behind the Stage😎🎥! Let me know your thoughts on this AI-accelerated VFX pipeline and how such VFX shots were previously achieved. Sony Pictures Entertainment #AIvfx #GaussianSplatting #Sony #Nvidia #NvidiaRTX #Jawset #Postshot #AItraining #AIcg #Compositing #VisualEffects #CG #AIcg Sony Electronics | Media Solutions (US & Canada)

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VFX, 3DGaussianSplatting, SonyVenice2, AIvfx, GaussianSplatting

• VR Digital Twin Radiance — Radiance Field & Volumetric 3DGS R&D

  Compositing two different #3D #GaussianSplatting layers on top of each other! All the loader truck, the ground, and the bottle are PhotoScanned 3D assets trained with #Postshot and composited in #AfterEffects. Let me know what you think! Trained and rendered on BTS | Behind the Stage Media servers. #AI #AIvfx #VFX #AIcg #AI3D #Nvidia #NvidiaRTX #Jawset #3DGS #Compositing #AIvisualeffects #Visualeffects

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Postshot, AfterEffects, AIvfx, VFX

• Volumetric 3DGS Digital Twin — Radiance Field & Volumetric 3DGS R&D

  What do you think? Was this kind of shot possible before #AI? #GaussianSplatting Thanks to new #AItools for #VFX artists. #AIvfx #AIcomposite #jawset #PostShot #AfterEffects #AImasking #Photoscanning #machinelearning #3dGS Rendered and Trained on BTS | Behind the Stage Media Servers.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, AItools, VFX, AIvfx, AIcomposite

• AI Generative 3DGS Training — Radiance Field & Volumetric 3DGS R&D

  Experimenting with this kind of realism in #VFX was not easily possible before the #AI era. I captured this shot of kids in the poolside with their mom's scarf and #iPhone in ProRes. The green bottle in the sand was photo-scanned and then trained using #PostShot #GaussianSplatting. I don't know what to say more!!! Let me know if you have any ideas. (Rendered and trained on BTS | Behind the Stage media server.) #AI #AItraining #Jawset #Aftereffects #AIvfx #VFX #CG #AIfilm #AI_Film_Making #VisualEffects #Splats #RTX #Nvidia #AImodel #Training #Adobe #Chromakey #MatchMove #Tracking #Composition #CameraTrack

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VFX, iPhone, PostShot, GaussianSplatting, AItraining

• AI Tools for VFX Production — Real-Time Tracking & VP Pipeline

  The Real set alignment test with the Digital set in #VirtualProduction.  Utilizing #RoboticArm tracking data seamlessly in #UE5.3 via live link, eliminating the need for separate optical tracking systems. Conducted at the Garage Studio, BTS | Behind the Stage, and DIGI TECH Trading L.L.C. Special thanks to my friend Ramez Tacrity for setting up the Zinema Motion arm. #Robots #Robotics #Tracking #UnrealEngine #XR #XRStudio #Xposure #Realtime #RealTimeVFX #VFX #CG #CGI #VP

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, RoboticArm, UE5, Robots, Robotics

• 3DGS Layer Compositing — Real-Time Tracking & VP Pipeline

  Here is the application of the mechanical tracking system and robotic arms in #VirtualProduction at Sharjah Xposure in Garage Studio and BTS | Behind the Stage. In collaboration with DIGI TECH Trading L.L.C and Zinema Motion. The tracking data is directly sent to #UnrealEngine 5.3 via LiveLinkFreeD! The environment is ElectricDreams, retouched by my good ue artist friends bharadwaj sriram and Shiva Tunoly. #tracking #trackingsystem #development #Innovation #UAE #sharjah #Dubai #XRstudio #VPstudio #VP #XR #robatics #robotarm #automation #event #exhibition

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, UnrealEngine, tracking, trackingsystem, development

• Sony Venice 2 Volumetric VFX — Real-Time Tracking & VP Pipeline

  Presenting the outcome of DMX Pixel Mapping development within Unreal Engine 5 during the Xposure 2024 Exhibition at Garage Studio and BTS | Behind the Stage #VirtualProduction Stage. Notice how adjustments in the lighting within the #UE5 editor seamlessly impact the set lights. #UnrealEngine #DMXPixelMapping #DMX #Gaffer #Lighting

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, UE5, UnrealEngine, DMXPixelMapping, DMX

• iPhone PostShot AI Pipeline — Real-Time Tracking & VP Pipeline

  🌟 Dive into the draft idea of #InteractiveLEDs, revolutionizing #Events with limitless 3D design and mesmerizing LED movements controlled by visitors' hands or #robotics all in #Realtime! The boundary is your imagination and #design prowess! Crafted with innovation at the Garage Studio and BTS | Behind the Stage laboratory. 🚀✨ #UE5 #UnrealEngine #InteractiveExperience #EventTech #LEDShowcase #DubaiEvents #ExhibitionDesign #RoboticInnovation #EventOrganizers #Motiondesign #3D #Interactive3D #GameEngine #Roboticarm #Realtime #RealtimeVFX #RealtimeCG #Interactive #InteractiveMedia #InteractiveSolutions

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: InteractiveLEDs, Events, robotics, Realtime, design

• Robotic Arm UE5 VP Stage — Real-Time Tracking & VP Pipeline

  Experimenting with #AR layers in #VirtualProduction at Garage Studio and BTS | Behind the Stage! 🎥🚀 Powered by native #UE5 and Pixotope tracking system for a seamless experience. 🌐✨ #RealtimeVFX #NvidiaRTX #XR #Dubai #XRstudio #UnrealEngine #VFX #CG #RealTimeCG #Composition

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, UE5, RealtimeVFX, NvidiaRTX, Dubai

• Unreal Engine Camera Tracking — Real-Time Tracking & VP Pipeline

  Exploring real-time #AR layers in virtual production at BTS | Behind the Stage and Garage Studio using native #UE 5.3! The possibilities are limitless, but challenges include synchronization across Render Nodes, Composite nodes, cameras, Tracking Systems, LED processors, and more. Using Pixotope tracking system. #RealtimeVFX #Nvidia #UnrealEngine #AI #VirtualProduction #XR #XRstudio #RealTime #VisualEffects #Compositing #RTX #UE5 #LEDvolume #Pixotope #TrackMen

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: RealtimeVFX, Nvidia, UnrealEngine, VirtualProduction, XRstudio

• UE5 Virtual Production Stage — Real-Time Tracking & VP Pipeline

  Experience a Bistro Restaurant in #VirtualProduction! Using the #SonyVenice2 and Pixotope at Garage Studio and BTS | Behind the Stage. Using Marketplace Bistro Restaurant Scene in #UE5.3! 🎥✨ #UnrealEngine #XR #Dubai #UAE #VP #nDisplay #EpicGames #MediaServer #NVIDIA #Cinematography

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, SonyVenice2, UE5, UnrealEngine, Dubai

• Real-Time AR Broadcast UE5 — Radiance Field & Volumetric 3DGS R&D

  Explore the face #motioncapture comparison between #UE 5.3 #Metahuman Animator operating in offline mode and #iClone 8 Live Face in real-time, within Sentient by Elysian and BTS | Behind the Stage Laboratory. I admire how Metahuman Animator generates the face mesh based on the video, while iClone impresses with its detailed motion capturing. Epic Games, Reallusion #aiAnimation #ImageProcessing #AI

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: motioncapture, Metahuman, iClone, aiAnimation, ImageProcessing

• Interactive LED Robotics Stage — Real-Time Tracking & VP Pipeline

  Testing DMX lighting control in UE 5.3 at Garage Studio and BTS | Behind the Stage, while evaluating tracking with ceiling-mounted light fixtures covering markers. 🎥💡 Godox Photo Equipment Co Ltd, Pixotope, Absen, #VirtualProduction #DMXControl #UE5 #Dubai #XR #DMX #Lighting #Arizona #LED #REDkomodo #UnrealEngine

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, DMXControl, UE5, Dubai, DMX

• Metahuman MoCap Live Link — Radiance Field & Volumetric 3DGS R&D

  Here is a glimpse of Gaussian Splatting as a 3D background in Virtual Production, featuring the new SONY BURANO at Garage Studio, in collaboration with BTS | Behind the Stage. A big thanks to Luma AI for their work in processing and training the splat, and seamlessly integrating it into Unreal Engine 5.3. #GaussianSplatting #Splat #VirtualProduction #VP #UE5 #UnrealEngine #PhotoScanning #Tracking #VFX #3D #Dubai #Filmmaking #Nvidia #nDisplay #LEDwall #Sony #SonyBurano #Burano #Pixotope #CG #onset #RealTimeVFX #RAW #ColorCorrection #ColorGrade #PhotoGrammetry #NeRF #AI #AItraining

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, Splat, VirtualProduction, UE5, UnrealEngine

• Sony Venice 2 Live VP Studio — Radiance Field & Volumetric 3DGS R&D

  Explore the implementation of Gaussian Splatting within UE 5.3 and new VFX workflow in SENTIENT BY ELYSIAN and BTS | Behind the Stage laboratory. Using Luma AI's new Interactive Scenes and updated UE plugin supporting Splats. The Depth Map is AI-generated and the splat is composited with the motion-captured #Metahuman character. #GaussianSplatting #VFX #AI #GenerativeAI #DepthMap #UE5 #UE #UnrealEngine #photoshop #Photogrammetry #LumaAI #Composition #3D #AIanimation #Animation #CG #Dubai #MotionCapture #ImageProcessing #CG

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Metahuman, GaussianSplatting, VFX, GenerativeAI, DepthMap

• DMX Stage Lighting UE5 Control — Real-Time Tracking & VP Pipeline

  Here is the Generative AI. 🚀 This video is AI-generated in Sentient by Elysian and BTS | Behind the Stage Laboratory. The goal of this project was to utilize AI tools to establish an approach that enables the swift creation of realistic video content with high quality, thereby meeting the market demands for expedited delivery in the Event and Film industries. All the animation, dialogue, voice-over, and music in this video are AI-generated. 🚨 #AI #GenerativeAI #AIGeneratedVideo #AImusic #AIimage #AIvideo #AItext #AIfilm #AIart #AIworkflow #AIanimation #AIdubai #Dubai #Dubaiai #AIplatform

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GenerativeAI, AIGeneratedVideo, AImusic, AIimage, AIvideo

• Real-Time VFX Raytracing — Real-Time Tracking & VP Pipeline

  Testing Customized #AR Solution with NVIDIA Omniverse in BTS | Behind the Stage. The capabilities of #AR paired with the raw power of a #Workstation #GPU with technologies like #DLSS, #Raytrace, and #Pathtrace for #RealTime #Rendering. Using #USDComposer software, an #iPhone, and an #RTX4090, with the ability to edit the 3D scene #live. 🌟 Huge shoutout to NVIDIA for providing the infrastructure that brings this Quality of AR experience to life. Just imagine the possibilities when you amp up your GPUs on your workstation! 🚀✨ #Omniverse #Nvidia #ARDeveloping #Developing #RTX #CG #VFX #RealtimeVFX #OnSet #Composite #ARtablet #Apple #iPhone14 #RealTimeRendering #Composition #RealTimeComposite #AugmentedReality

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Workstation, GPU, DLSS, Raytrace, Pathtrace

• Gaussian Splatting VP Stage — Radiance Field & Volumetric 3DGS R&D

  Presenting the 3D #GaussianSplatting of 200 #helishot images in BTS | Behind the Stage, capturing the exquisite Tomb of Hafez in #Shiraz, #Iran. Shot in 2019 and seamlessly processed today by Polycam in just 20 minutes. This is part of a collection of 5675 images taken for the heritage application of #Photogrammetry in #VirtualReality using #RealityCapture. Experience this immersive #3DGaussian journey now through the link below: https://lnkd.in/dpuCdGeJ #3DReconstruction #Photoscanning #NeRF

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: GaussianSplatting, helishot, Shiraz, Iran, Photogrammetry

• Metahuman 3DGS VFX Pipeline — Real-Time Tracking & VP Pipeline

  Here is the journey in the Garage Studio and BTS | Behind the Stage to achieve precise tracking data using the Pixotope tracking system within #UnrealEngine. 🎥🔍 Our ultimate aim? Seamlessly synchronizing #AR Layers in #RealTime with the camera footage. 🌟 #VirtualProduction #TimeCode #TimeStep #UE5 #Tracking #OpticalTracking #LiveLink #VFX #RealTimevfx #Composition

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: UnrealEngine, RealTime, VirtualProduction, TimeCode, TimeStep

• Generative AI Video Workflow — Radiance Field & Volumetric 3DGS R&D

  Exploring and testing the magic of #MotionCapture through #ImageProcessing based on an #iPhone camera wifi feed in #RealTime at BTS | Behind the Stage. Powered by NVIDIA Omniverse, RTX #AI processing. 🌟 #Dubai #CG #RealtimeVFX #GenerativeAI

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: MotionCapture, ImageProcessing, iPhone, RealTime, Dubai

• GPU DLSS Raytracing Rig — Radiance Field & Volumetric 3DGS R&D

  Presenting a glimpse into the world of 2.5D technique (Camera Mapping) in #VirtualProduction in BTS | Behind the Stage and Garage Studio. This 3D landscape draws inspiration from a photograph captured at the Laxenburg Catholic Church in Austria. The transformation journey involved a blend of #Photoshop, #Blender, and #UnrealEngine - that graced the LED wall. Powered by the Pixotope tracking system. Music: Gole Yakh #VirtualProduction #Austria #Dubai

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, Photoshop, Blender, UnrealEngine, Austria

• Blender to UE5 VP Workflow — Radiance Field & Volumetric 3DGS R&D

  Exciting car shoot test in Garage Studio and BTS | Behind the Stage! 🚗🎥 A big shoutout to my lovely Filipina friends for their incredible help in making this happen. The background is a video plate captured with my mobile while cycling, brought to life with #DavinciResolve on the LED Wall. #Video #Dubai #Car #Studio #Filmmaking #VirtualProduction

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: DavinciResolve, Video, Dubai, Car, Studio

• Shiraz Helishot Aerial Splat — Radiance Field & Volumetric 3DGS R&D

  Unlock the extraordinary world of Johanna's Digital Twin in BTS | Behind the Stage. Dubsmashing the tune of old Iranian singers, Lady Delkash and Vigen! 🎙🎻🎵 Her face crafted from photo scanning with #realitycapture, powered by #Metahuman and #UnrealEngine 5.2.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: realitycapture, Metahuman, UnrealEngine, metahuman, unrealengine

• Real-Time AR in Unreal Engine — Radiance Field & Volumetric 3DGS R&D

  Excited to share a behind-the-scenes look at my adventures in Dubai! 🌟 Watch as I capture footage of #DubaiExpo, #DubaiMarina, and the iconic #BurjKhalifa while riding my bicycle and public scooter. These locations serve as the perfect backdrop for the #HarleyDavidson bike in BTS | Behind the Stage and Garage Studio. #VirtualProduction #VideoPlates #DavinciResolve

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: DubaiExpo, DubaiMarina, BurjKhalifa, HarleyDavidson, VirtualProduction

• Metahuman Photogrammetry UE5 — Radiance Field & Volumetric 3DGS R&D

  Exploring #Nerf in #VirtualProduction with real-time #AR layers powered by #UnrealEngine 5.2 at Garage Studio with the help of Luma AI. Appreciation to BTS | Behind the Stage for their support in bringing these developments to life. The 3D Background is Museum of Future in #Dubai. #XR #development

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Nerf, VirtualProduction, UnrealEngine, Dubai, development

• iPad MoCap Image Processing — Real-Time Tracking & VP Pipeline

  Zoom out #Dubai and explore the Earth using #UnrealEngine 5.2, powered by the innovative Cesium solution. The potential is limitless, spanning from simulation applications to film and media content creation. Special thanks to BTS | Behind the Stage for their support in driving these advancements.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: Dubai, UnrealEngine, dubai, unrealengine

• DaVinci Resolve VP Color — Real-Time Tracking & VP Pipeline

  I am excited to unveil the ongoing development process of SET EXTENSION in the Garage Studio's #VirtualProduction. 🎥 Experience the real-time compositing in action, powered by #UnrealEngine 5.2 and the Pixotope tracking system. Witness the seamless integration of CG foreground and background elements with the live camera feed, as they come together in perfect harmony. Expressing deep appreciation to Sattar K. for the exceptional 3D environment created! A special thanks to BTS | Behind the Stage for their invaluable support in driving these developments. #XR #AR #SetExtension #VisualEffects #VirtualStudio

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, UnrealEngine, SetExtension, VisualEffects, VirtualStudio

• Burj Khalifa Dubai Marina Scan — Radiance Field & Volumetric 3DGS R&D

  Introducing the Remote Control for #UnrealEngine at Garage Studio, now accessible through an #iPad on the set! This innovative development empowers #Cinematographers to effortlessly modify stage positions, colors, light cards, and other parameters wirelessly, enabling them to create the perfect frame. A huge shoutout to BTS | Behind the Stage for their invaluable assistance and unwavering support in bringing these advancements to life. #VirtualProduction #FilmMaking #Innovation

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: UnrealEngine, iPad, Cinematographers, VirtualProduction, FilmMaking

• NeRF Virtual Production AR — Radiance Field & Volumetric 3DGS R&D

  Check out this development of using the #MetaQuest2 controller as a tracker with #UnrealEngine! With this setup, the camera position inside UE is seamlessly linked to the controller, unlocking endless possibilities. Witness the stunning #Hillside UE sample project in action. Special thanks to BTS | Behind the Stage for their invaluable help and support in bringing these advancements to life. #Democratization #Technology #VirtualProduction #MetaQuest

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: MetaQuest2, UnrealEngine, Hillside, Democratization, Technology

• Unreal Engine Dubai Showcase — Radiance Field & Volumetric 3DGS R&D

  Join me as we have a playful exploration of technology and uncover its boundless potential! 🎮🚀💡😜🔬 Introducing a #VirtualProduction experiment conducted at home. 🏠✨ Witness the prototype that leverages an iPhone and iPad as a camera, tracking system, and real-time facial motion capture device, all in one. 📱💻🎥 Behold the magic of an #AR production powered by #UnrealEngine and #MetaHuman. 🌟 Enormous gratitude to my dear friends Bartosz Jankowski and Joanna for their invaluable assistance in making this test a reality! 🙏🤝 Special thanks to BTS | Behind the Stage for their incredible help and support! 🌟🙌 #democratization #technology

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, UnrealEngine, MetaHuman, democratization, technology

• Metahuman CG Characters UE5 — Real-Time Tracking & VP Pipeline

  Presenting a development test of #MetaHuman in #UnrealEngine 5.2, showcasing its incredible capabilities to dubsmash Lady #Mahasti, one of Iran's most renowned singers, using #FaceLiveLink. A heartfelt thank you to BTS | Behind the Stage for their invaluable assistance in bringing these exciting advancements to life. #CGDevelopment #RealisticAnimation #VirtualDubsmash

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: MetaHuman, UnrealEngine, Mahasti, FaceLiveLink, CGDevelopment

• iPad Virtual Camera Setup — Radiance Field & Volumetric 3DGS R&D

  Witness the astounding ability to create lifelike CG characters that closely resemble their real-life counterparts. On the left, we have a Photo Scanned Model, while on the right, behold the incredible #MetaHuman based on this very model. Special thanks to BTS | Behind the Stage for their invaluable assistance and support in bringing these developments to life. Unlock the power of #MetaHuman and harness the capabilities of Unreal Engine, courtesy of Epic Games. #CGCharacters #Realism #UnrealEngine #EpicGames #BehindTheStage

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: MetaHuman, CGCharacters, Realism, UnrealEngine, EpicGames

• Unreal Engine AR Broadcast — Real-Time Tracking & VP Pipeline

  Exciting developments in the #VirtualCamera realm! Testing and refining a Virtual Camera within Unreal Engine 5.2 in the mesmerizing Electric Dreams Environment has been an incredible experience. It's like operating a drone, but without the fear of collisions or signal loss. Plus, the added bonus of enjoying the cool AC while doing it! 😋 Thanks to the support of BTS | Behind the Stage, we've been able to achieve remote control of camera movement and specifications through an iPad. The possibilities are truly limitless! #VirtualCamera #UnrealEngine5.2 #VirtualProduction

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualCamera, UnrealEngine5, VirtualProduction, virtualcamera, unrealengine5

• Meta Quest 2 Hillside Setup — Radiance Field & Volumetric 3DGS R&D

  I'm thrilled to announce the exciting progress we've made in the Garage Studio's #virtualproduction! With the invaluable assistance of BTS | Behind the Stage, we have successfully integrated CG foreground layers into our workflow. One notable addition is the AR Foreground "No Parking Sign", which not only captures attention but also seamlessly transitions the focus from this AR SIGN to the REAL BMW BIKE. I would like to extend a special thanks to environment artist Sattar K. for his outstanding work in creating this captivating scene using #unrealengine5. Our tracking system of choice is Pixotope, while all the Wall and AR CG elements have been developed using UE5.2. Join us in exploring the future of #VP and unlocking the full potential of #VirtualProduction, #AR, and #AugmentedReality with #UE5 and #UnrealEngine.

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: virtualproduction, unrealengine5, VirtualProduction, AugmentedReality, UE5

• Virtual Camera UE5 Pipeline — Radiance Field & Volumetric 3DGS R&D

  Using #NeRF in #VirtualProduction : Check out this video showcasing the development with NeRF in Garage Studio. It marks the beginning of an incredible journey, made possible with the invaluable help and support of BTS | Behind the Stage. Special thanks to the remarkable HTC VIVE MARS tracking system for its useful contributions. #AI #NeRF #UnrealEngine #VirtualProduction #Innovation

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: NeRF, VirtualProduction, UnrealEngine, Innovation, nerf

• Mahasti Metahuman Digital Twin — Radiance Field & Volumetric 3DGS R&D

  Showcasing the use of 3D NeRF as backgrounds in Unreal Engine 5.2 for VFX purposes.📸📹 In the behind-the-scenes footage, you'll witness two scenes that have been photoscanned and then reconstructed using NeRF technology, all thanks to Lumalabs.ai. Special thanks to BTS | Behind the Stage for their incredible help and support in making these developments possible. I couldn't have done it without you! 🙌🏼 #VFX #UnrealEngine #NeRFtechnology #Lumalabs #ai #Dubai #nerf #lumalabs

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VFX, UnrealEngine, NeRFtechnology, Lumalabs, Dubai

• XR Stage VP Production — Radiance Field & Volumetric 3DGS R&D

  "Excited to showcase my first #NeRF test: Cooking with LumaLabs.ai! 🍳🔥" link to check out this NeRF: https://shorturl.at/irwzQ #ai #nerf #3dmodeling #nvidia

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: NeRF, nerf, 3dmodeling, nvidia

• Unreal Engine 5 VP Pipeline — Real-Time Tracking & VP Pipeline

  "Presenting the exciting Virtual Production Development journey of Garage Studio and BTS | Behind the Stage with the remarkable Pixotope Tracking system. 🎬✨ #VirtualProduction #Dubai #Development #xrstudio

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: VirtualProduction, Dubai, Development, xrstudio, virtualproduction

• NeRF AI 3D Modeling — Radiance Field & Volumetric 3DGS R&D

  One of the cool specialties of Unreal Engine and Virtual Production is Tracking your light fixtures on the set and aligning them with the corresponding light in the digital scene that is getting rendered on the LED wall, all live on the set. It's Fractal Studio the Pioneer in XR applications in the film industry in Dubai. #film #dubai #vp #xr #ar #virtualproduction #unrealengine #newmedia #uae #cg #visualeffects #vfx #xr_developer #vp_developer

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: film, dubai, virtualproduction, unrealengine, newmedia

• Mo-Sys Unreal VP Pipeline — Real-Time Tracking & VP Pipeline

  My first dialogue with #chatgpt I asked my concern these days, "how to increase the fps in the unreal engine?" And it started to reply... As is shown in the video in real-time. The answer is INSANE! I didn't expect such details. Pixojam Virtual Production #ai #machinelearning

  Challenge: Synthesizing and calibrating multi-view spatial datasets with high fidelity.

  Result: Production-grade asset synthesis with verified real-time performance.

  Technologies: chatgpt, machinelearning

• Mo-Sys Metahuman Real-Time AR — Real-Time Tracking & VP Pipeline

  This is an AR test. In real-time, the camera position is tracked and both the fire and 3d guy (meta-human) as CG layer are composited in real-time with unreal engine 5.1. Pixojam Virtual Production #mosys #unrealengine #virtualproduction #augmentedreality

  Challenge: Real-time camera tracking and live CG compositing of dynamic particle simulation and MetaHuman characters.

  Result: Zero-latency real-time spatial sync and in-camera visual effects composite.

  Technologies: Mo-Sys, Unreal Engine 5.1, MetaHuman, Real-Time AR, Virtual Production