5 papers
MMSI-Bench: A Benchmark for Multi-Image Spatial Intelligence
Sihan Yang, Runsen Xu, Yiman Xie +10
Spatial intelligence is essential for multimodal large language models (MLLMs) operating in the complex physical world. Existing benchmarks, however, probe only single-image relati…
UltraDexGrasp: Learning Universal Dexterous Grasping for Bimanual Robots with Synthetic Data
Sizhe Yang, Yiman Xie, Zhixuan Liang +4
Grasping is a fundamental capability for robots to interact with the physical world. Humans, equipped with two hands, autonomously select appropriate grasp strategies based on the…
InternVLA-A1: Unifying Understanding, Generation and Action for Robotic Manipulation
Junhao Cai, Zetao Cai, Jiafei Cao +39
Prevalent Vision-Language-Action (VLA) models are typically built upon Multimodal Large Language Models (MLLMs) and demonstrate exceptional proficiency in semantic understanding, b…
MMSI-Video-Bench: A Holistic Benchmark for Video-Based Spatial Intelligence
Jingli Lin, Runsen Xu, Shaohao Zhu +11
Spatial understanding over continuous visual input is crucial for MLLMs to evolve into general-purpose assistants in physical environments. Yet there is still no comprehensive benc…
InternData-A1: Pioneering High-Fidelity Synthetic Data for Pre-training Generalist Policy
Yang Tian, Yuyin Yang, Yiman Xie +13
Recent works explore how real and synthetic data contribute to Vision-Language-Action (VLA) models' generalization. While current VLA models have shown the strong effectiveness of…