ConsisVLA-4D: Advancing Spatiotemporal Consistency in Efficient 3D-Perception and 4D-Reasoning for Robotic Manipulation
arXiv:2605.05126v1 Announce Type: new
Abstract: Current Vision-Language-Action (VLA) models primarily focus on mapping 2D observations to actions, but exhibit notable limitations in spatiotemporal perception and reasoning: 1) spatial representations o…