HKU and VAST's Mira-Scene fixes object placement in single-image 3D scene reconstruction
jiqizhixin · x · 2026-10-08
Single-image 3D generation has mastered object-level reconstruction, but scaling to full scenes introduces a placement problem: generated objects look right yet never quite line up with the original image.
The University of Hong Kong and VAST present Mira-Scene, which introduces a pixel-aligned layout representation: it first derives a layout aligned to the source image at the pixel level, then conditions 3D generation on that layout, guaranteeing strict consistency between the reconstructed scene and the input. It can be combined with GPT-6 Astra.
More from Multimodal
- Person Remover LoRA for MiniMax H3 Released with Full ComfyUI Workflow — linoy_tsaban · 2026-10-08
- KAIST's Tetris3D Reconstructs 3D Scenes With Physically Coherent Objects, Plus 1.2M-Scene ComOb Dataset — kaist-ai · 2026-10-08
- Suno Praised as Good Enough to Eventually Overtake Spotify — petergyang · 2026-10-08
- 3.5M-param anime upscaler applies Looped-DiT recurrence, gains +0.6dB in 4 loops — NobodySnJake · 2026-10-08
- Musk Hypes Grok Bot After User Generates Full Explainer Video From a 15-Second Prompt — elonmusk · 2026-10-08
- AI pop singer Claudia open-sourced: downloadable soul via MCP server, skills and character sheets — Promptmethus · 2026-10-08