This paper studies category-level RGB-D object pose estimation, which recovers an object's 3D rotation and translation without instance-specific CAD models or reference views. This task is highly ambiguous due to intra-class variations, symmetries, and partial observations. Current solutions struggle to balance reliabi...
Jin-Yu Zhang, Haitao Lin, Jia-Shu Hou et al.· 0 citations
FineVLA, an open framework for action-aligned fine-grained VLA supervision, is introduced and the largest real-world gains appear on pose, color, and approach direction, and approach direction--factors where goal-level instructions provide no guidance.
Across in-domain and distribution-shifted settings, allocating additional test-time computation substantially improves next-subtask prediction accuracy, and these gains translate into higher closed-loop success on long-horizon robot manipulation tasks.