GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives

原文：GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives

作者：Qize Yu, Lianrui Fan, Boyu Chen, Jiaqi Liang, Xini Ding, Yue Chen, Zetian Song, Yuran Wang, Yi Zou, Kaixuan Wang, Tianxing Chen, Wenxuan Song, Bohan Zhou, Mingleyang Li, Siqiao Huang, Yuqi Ye, Caigao Jiang, Wei Wei, Ruihai Wu, Hang Zhang, Yixiao Ge, Shuchang Zhou, Shilong Liu, Xianming Liu, Ping Luo

来源：arXiv cs.RO（机器人）

正文

Computer Science > Computer Vision and Pattern Recognition

arXiv:2609.39601v1 (cs)

[Submitted on 30 Sep 2026]

Title:GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives

Authors:Qize Yu, Lianrui Fan, Boyu Chen, Jiaqi Liang, Xini Ding, Yue Chen, Zetian Song, Yuran Wang, Yi Zou, Kaixuan Wang, Tianxing Chen, Wenxuan Song, Bohan Zhou, Mingleyang Li, Siqiao Huang, Yuqi Ye, Caigao Jiang, Wei Wei, Ruihai Wu, Hang Zhang, Yixiao Ge, Shuchang Zhou, Shilong Liu, Xianming Liu, Ping Luo, Shiyu Huang

View a PDF of the paper titled GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives, by Qize Yu and 25 other authors

View PDF

HTML (experimental)

Abstract:Precise grounding matters. It specifies which object is the target and where that object is, even in clutter and for tiny objects, and it has to be fast enough for closed-loop control. Yet vision-language-action (VLA) and world-action models (WAMs) take perception from general-purpose vision-language and video-generation backbones, which still fail in these settings. We introduce GroundingPI, a 4B grounding foundation model that generates points and boxes as quantized coordinates in a shared vocabulary. Training combines multimodal and spatial pretraining, supervised fine-tuning, and reinforcement learning with GRPO, using supervision from public datasets and dedicated data engines. Against 44 baselines across 34 grounding benchmarks spanning 11 perceptual capabilities, GroundingPI establishes a new state of the art, averaging 73.68%, above the larger GPT-6 Astra (71.54%). As a downstream visual backbone, GroundingPI improves performance on robotic manipulation and autonomous driving. On RoboTwin 2.0, it outperforms every mainstream backbone we evaluate in all four out-of-distribution settings, by up to 24.8% relative to the strongest backbone. On RoboCasa-GR1, GroundingPI trained with 50% of the demonstrations outperforms those baselines trained with 75%. On nuScenes, used as the visual backbone, GroundingPI attains an average open-loop L2 error of 0.296 m. We systematically analyze GroundingPI’s pretraining in scale and data composition. Downstream autonomous driving and robotic manipulation improve as the pretraining is scaled. Analyzing the data recipe across these 11 perceptual capabilities shows dense grounding’s substantial benefits for both, and OCR’s potential as a catalyst for perceptual learning. These results support grounding as a perceptual foundation, and dedicated perceptual pretraining as a promising direction for foundation models of physical intelligence.

Comments:

64 pages, including supplementary material. Project page: this https URL Code: this https URL Model: this https URL

Subjects:

Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Robotics (cs.RO)

ACM classes:

I.2.10; I.2.6; I.2.9

Cite as:

arXiv:2609.39601 [cs.CV]

(or

arXiv:2609.39601v1 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2609.39601

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

主题

计算机视觉 · 人工智能 · 机器学习 · 机器人

由「前沿雷达」于 2026-09-30 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。