Open-source HARMONY uses VLM agents to reconstruct 3D scenes, matching GPT-6 Astra quality
thoma_gu · x · 2026-09-24
The HARMONY team released an open-source framework that reconstructs compositional 3D scenes from a single indoor image using a VLM agent. Like a 3D artist, it builds hierarchically — room structure and camera first, then wall-mounted objects, furniture, and decorations — comparing renders to the input and iteratively correcting errors. Even with older VLM backbones it is competitive with GPT-6 Astra in visual quality with better geometric alignment, alongside the new HARMONY300 benchmark.
Related event: UPenn Open-Sources HARMONY for Single-Image 3D Scene Reconstruction(3 posts)→
More from Research
- Self-organizing agent teams average 66.7% vs 48.8% for best single model, Stanford and Together AI paper finds — james_y_zou · 2026-09-24
- OpenRSI-Index v0.1 launches to benchmark recursive AI self-improvement on 1k-GPU clusters — ChengleiSi · 2026-09-24
- New paper on adversarial delegation: agent picked a $601 flight over a $91 one after reading your emails — niloofar_mire · 2026-09-24
- SchrödingerRepo: rewriting repos exposes LLM memorization on SWE-bench — SJTU · 2026-09-24
- Survey: memory mechanisms for autoregressive video generation — Harold Haodong Chen · 2026-09-24
- Robotics' real frontier is post-training: self-play in sim closes the demo-to-deployment gap — ZGojcic · 2026-09-24