spatial-boost

【设计与创意】Inject spatial understanding into VLMs via language-guided multi-turn Chain-of-Thought reasoning over hierarchical spatial knowledge (pixel→object→scene). Implement dual-channel attention mechanism preserving pre-trained vision features while progressively incorporating dense 3D spatial information. Frozen original parameters prevent catastrophic for…

独立 GitHub 快照 · 自动分类 · 以原作者文档为准