BVB benchmark tests agent video understanding by rebuilding 288 real videos in Blender
SonglinYang4 · x · 2026-09-15
- Researchers from University of Rochester, Sony, CMU and UW introduce BVB (Blender-VideoBench), premised on the idea that a truly video-understanding agent should be able to reconstruct videos programmatically: agents must code animated Blender scene reconstructions of real footage, replacing QA-style evaluation.
- The benchmark covers 288 real videos across scenes like ARKitScenes and ScanNet++ and tests 51 agent configurations, with all agents running a lightweight harness (Mini-BVB) in identical sandboxes under shared cost limits. Paper, code and leaderboard are public.
Related event: BVB Benchmark Tests Video Understanding via Blender Reconstruction(2 posts)→
More from Multimodal
- Audio8-TTS-Realistic-Air, a zero-shot voice-cloning TTS model, trends on Hugging Face — Edge0 · 2026-11-02
- Mistral explains why its model-built solver is a first step toward new drugs and batteries — MistralAI · 2026-11-02
- Mistral Large 4 writes a quantum chemistry SCF solver from scratch with one short prompt — MistralAI · 2026-11-02
- Gorgeous dithered pixel art collage serves up a 'great morning' — philipvollet · 2026-11-02
- Clown Prophecy moves to new repo, now uses first clip as reference to fix H3 quality drift — reeight · 2026-11-02
- A 25-year design veteran asks: who still mixes pro software with AI tools? — Opening-Knee-5913 · 2026-11-02