The proposed exploration-augmented key-observation gating first determines whether the current RL episode is judgment-worthy based on its similarity to a generated success trajectory, and proposes VLM success reward via Exploration-Augmented Key-Observation Gating (VLMeg) to well harness VLM for embodied reward design.
Xin Liu, Yu-Tong Jiang, Jia-Yue Su et al.· 0 citations
We present ContinuousCloth, a continuous-time, physics-informed neural solver for deformable cloth dynamics that enforces a spring--mass ODE within a neural trajectory model. Unlike classical time-stepping simulators that return trajectories only on a discrete integration grid, ContinuousCloth supports querying at arbi...
Yingdong Ru, Lipeng Zhuang, Zhuo He et al.· 0 citations
Humans and animals do not learn vision from a fixed set of data: they move, interact, and learn from the visual consequences of their actions. Motivated by this developmental view, we study whether robots can improve visual learning by actively collecting experience through curiosity-driven interaction. We introduce Dy...
Stelios Zarifis, P. Filntisis, George Retsinas et al.· 0 citations
Any-angle path planning extends traditional graph-based path planning by allowing movement between any pair of vertices, rather than being restricted by predefined edges. It can find straighter and shorter paths in continuous space with graphs, making it particularly suitable for navigation in open areas. Many any-angl...
Yiyuan Zou, Clark Borst· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Real-world evaluation is the gold standard for robot manipulation policies because it tests them against the physical conditions and deployment challenges they are ultimately designed to handle. However, real-world evaluation is also the bottleneck for iterating on robot imitation learning policies: it is costly, diffi...
Haoxu Huang, Tongsam Zheng, Yifan Chen et al.· 0 citations
Embodied foundation models have recently been widely used to improve robot generalization and task success rates. Previous works apply lossy efficient-inference techniques such as quantization, pruning, and asynchronous inference, accepting small action quality degradation in exchange for lower per-inference computatio...
Yujin Wang, Junli Chen, Yixuan Li et al.· 0 citations
Robot world models now generate synthetic manipulation videos, but evaluating them is hard: realistic-looking outputs often violate physical laws, temporal consistency or task logic, while conventional metrics and monolithic Vision-Language Model (VLM) judges return scores that do not say what failed. We present RoboGa...
Minh-Loi Nguyen, Nghiem T. Diep, Hung K. Nguyen et al.· 0 citations
Reinforcement learning for robot manipulation is often bottlenecked by reward design, especially in long-horizon tasks: sparse success rewards provide weak supervision, while hand-crafted dense rewards are tedious to design and generalize poorly across tasks. Progress-based reward models offer a promising alternative b...
Pengzhi Yang, Xinyu Wang, Pengyu Jing et al.· 0 citations
Vision-language models (VLMs) are increasingly used for scene understanding in autonomous driving, but robustness analysis often relies on task-agnostic embedding stability alone. We study whether corruption-induced embedding drift predicts changes in a task-aligned hazard score derived from CLIP image-text similaritie...
Deformable object manipulation (DOM) is challenging due to high-dimensional, partially observable states that evolve through long-horizon, topology-changing interactions with multiple valid manipulation modes. We introduce Instant-Fold, an in-context imitation learning framework for DOM. Given a single human demonstrat...
Yilong Wang, Cheng Qian, Edward Johns· 0 citations
We propose to turn generalist multi-embodiment value functions into reusable models for robot design. Instead of running a new reinforcement learning co-design loop for each robot, we first train an embodiment-aware policy and value function across many robot designs. After training, the frozen value function is used a...
This work introduces a VLA tool-family interface that exposes explicit tool selection and in-execution progress feedback, enabling efficient event-triggered agent replanning without continuous agent polling and proposes Tool-Aligned Post-Training (TAPT), which constructs invocation-aligned training units for instructio...
A new system identifies and corrects unreliable object descriptions stored in a robot’s memory, improving the robot’s ability to interact with its environment.
Robots are getting smarter, but how can their hardware match that growth? New Microsoft Research findings show that moving AI inference beyond the robot can improve task success, boost efficiency, and support more advanced physical AI workloads. The post Offloaded inference for real-world physical AI robotics appeared first on Microsoft Research.
