Track2Art is a motion‑centric framework that recovers articulated object models from RGB‑D interaction videos by lifting 2D point tracks into 3D trajectories. It groups these trajectories into rigid‑part hypotheses and uses learned‑analytic reasoning to infer directed kinematic relations, joint types, and joint geometry. On the PartNet‑Mobility benchmark, it achieves 0.695 Point IoU and 0.410 end‑to‑end J@20 without requiring ground‑truth part counts or test‑time optimization.
By Xiaotong Li, Yixiong Jing, Junsheng Ding, Weihang Li, Benjamin Busam, Guangming Wang, Brian Sheil
arXiv:2610.01744v1 Announce Type: new
Abstract: Robot manipulation models primarily reason from 2D observations while acting in the 3D physical world. To bridge this gap, recent work has augmented ro...
By Wonguen Cho, Junhoo Lee, Nojun Kwak
arXiv:2609.27675v2 Announce Type: replace
Abstract: Understanding articulated objects is fundamental for robotic interaction, requiring accurate rigid-part discovery and the recovery of their kinemat...
By Xiaotong Li, Yixiong Jing, Junsheng Ding, Weihang Li, Benjamin Busam, Guangming Wang, Brian Sheil
World models endow perceptual systems with the ability to predict how scenes evolve under interaction. They are most beneficial when trained on diverse volumes of data, to instill a rich prior into do...
arXiv:2609.19142v1 Announce Type: new
Abstract: World models endow perceptual systems with the ability to predict how scenes evolve under interaction. They are most beneficial when trained on diverse...
By Bardienus P. Duisterhof, Kaifeng Zhang, Adam Hung, Bowen Wen, Stan Birchfield, Yunzhu Li, Deva Ramanan, Jeffrey Ichnowski
arXiv:2607. 11167v1 Announce Type: cross Abstract: Representing manipulation actions as 2D trajectories in the camera plane provides a compact and interpretable basis for learning complex 3D manipulation policies.
By Haojie Huang, Linfeng Zhao, Haotian Liu, Zhang Ye, Si-Yuan Huang, Mingxi Jia, Boce Hu, Fangzhou Lin, Yu Qi, Dian Wang, Robin Walters, Robert Platt
arXiv:2609.01059v1 Announce Type: new
Abstract: As Vision-Language Models (VLMs) tackle dynamic 3D spatial reasoning, ego-motion perception becomes essential to resolve monocular scale ambiguity. How...
By Jiayu Ding, Zhuodong Liu, Lei Zhang, Manyu Xiong, Hongbo Jin, Haoran Tang, Hongbo Zhang, Changen Zhu, Wenbo Xing
arXiv:2609.19119v1 Announce Type: new
Abstract: Human videos contain rich causal evidence for robot manipulation: they reveal how hand motion induces object motion and produces task-relevant changes...
By Jiaming Zhang, Homanga Bharadhwaj
arXiv:2609.36940v1 Announce Type: new
Abstract: Accurate dynamic scene reconstruction is important for robotic perception, where temporally consistent representations of dynamic environments are esse...
By Thai Duy Nguyen, Haitian Zhang, Addison Lin Wang
Perception-based humanoid loco-manipulation requires connecting egocentric observations and task instructions to whole-body motion. Learning this mapping requires synchronized egocentric images, language commands, and robot-compatible kinematic trajectories, yet no existing data source provides this complete tuple at scale.
Grasp in Gaussians (GraG) is a fast, robust method for reconstructing dynamic 3D hand‑object interactions from a single monocular video. It leverages pretrained hand and object priors and represents the scene with a compact Sum‑of‑Gaussians (SoG) model, enabling efficient tracking while preserving geometric fidelity. Experiments show GraG achieves temporally coherent reconstructions on long sequences 4.4×–38.9× faster than prior work.
By Ayce Idil Aytekin, Xu Chen, Zhengyang Shen, Thabo Beeler, Helge Rhodin, Rishabh Dabral, Christian Theobalt
InfiNoVA is a data‑augmentation framework that transforms synchronized multi‑camera demonstrations into a dense, geometrically consistent set of training views by reconstructing each manipulation trajectory as a time‑varying 3D Gaussian. The method renders novel observations from sampled camera poses while preserving the original state‑action pairs, improving frame‑level fidelity and temporal consistency compared to generative synthesis. Across four real‑world manipulation tasks, policies trained with InfiNoVA achieve 5.4× higher average success under unseen randomized viewpoints than VISTA‑based augmentation and 1.7× higher success than training on all five physical camera views.
By Sai Puneeth Reddy Gottam, Elmar Rueckert, Vedant Dave