Paper Detail
Yuxuan Tian, Yurun Jin, Bin Yu, Yukun Shi, Hao Wu, Chi Harold Liu, Kai Chen, Cong Huang
Robotic manipulation critically requires reasoning about future spatial-temporal interactions, yet existing VLA policies and world-model-enhanced policies do not fully model action-relevant spatial-temporal interaction structure. We propose STARRY, a world-model-enhanced action-generation policy that aligns spatial-temporal prediction with action generation. STARRY jointly denoises future spatial-temporal latents and action sequences, and introduces Geometry-Aware Selective Attention Modulation to convert predicted depth and end-effector geometry into token-aligned weights for selective action-attention modulation. On RoboTwin 2.0, STARRY achieves 93.82% / 93.30% average success under Clean and Randomized settings. Real-world experiments further improve average success from 42.5% to 70.8% over $π_{0.5}$, demonstrating the effectiveness of action-centric spatial-temporal world modeling for spatial-temporally demanding robotic action generation.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@article{tian2026starry,
title = {STARRY: Spatial-Temporal Action-Centric World Modeling for Robotic Manipulation},
author = {Yuxuan Tian and Yurun Jin and Bin Yu and Yukun Shi and Hao Wu and Chi Harold Liu and Kai Chen and Cong Huang},
year = {2026},
abstract = {Robotic manipulation critically requires reasoning about future spatial-temporal interactions, yet existing VLA policies and world-model-enhanced policies do not fully model action-relevant spatial-temporal interaction structure. We propose STARRY, a world-model-enhanced action-generation policy that aligns spatial-temporal prediction with action generation. STARRY jointly denoises future spatial-temporal latents and action sequences, and introduces Geometry-Aware Selective Attention Modulation },
url = {https://arxiv.org/abs/2604.26848},
keywords = {cs.RO},
eprint = {2604.26848},
archiveprefix = {arXiv},
}
{}