7 papers
Dense Metric Depth Completion from Sparse Direct Time-of-Flight Sensors
Hakyeong Kim, Ruicheng Wang, Chengtang Yao +2
Direct Time-of-Flight (dToF) sensors provide highly accurate metric depth and are more robust than indirect ToF systems in challenging real-world conditions. However, their high ma…
Continue or Replan? Bernoulli-Continuation Policy Learning for Adaptive Horizon Execution
Weichen Xu, Zhenhua Liu, Lin Luo +8
Existing chunk-based Vision-Language-Action (VLA) models execute a fixed number of actions (i.e., execution horizon) before replanning, turning replanning into a task-agnostic peri…
MoGe-3: Fine-Detail Monocular Geometry Estimation with Self-Guided Sparse Volumetric Refinement
Lingyu Kong, Ruicheng Li, Ruicheng Wang +4
Monocular geometry estimation has recently achieved impressive performance across diverse scenes. However, state-of-the-art models still face notable distortion in local 3D structu…
Scalable Vision-Language-Action Model Pretraining for Robotic Manipulation with Real-Life Human Activity Videos
Qixiu Li, Yu Deng, Yaobo Liang +14
This paper presents a novel approach for pretraining robotic manipulation Vision-Language-Action (VLA) models using a large corpus of unscripted real-life video recordings of human…
3D Visual Illusion Depth Estimation
Chengtang Yao, Zhidan Liu, Jiaxi Zeng +3
3D visual illusion is a perceptual phenomenon where a two-dimensional plane is manipulated to simulate three-dimensional spatial relationships, making a flat artwork or object look…
Diving into the Fusion of Monocular Priors for Generalized Stereo Matching
Chengtang Yao, Lidong Yu, Zhidan Liu +3
The matching formulation makes it naturally hard for the stereo matching to handle ill-posed regions like occlusions and non-Lambertian surfaces. Fusing monocular priors has been p…