Phi-Ground Tech Report: Advancing Perception in GUI Grounding
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866908473988481024 |
|---|---|
| author | Zhang, Miaosen Xu, Ziqiang Zhu, Jialiang Dai, Qi Qiu, Kai Yang, Yifan Luo, Chong Chen, Tianyi Wagle, Justin Franklin, Tim Guo, Baining |
| author_facet | Zhang, Miaosen Xu, Ziqiang Zhu, Jialiang Dai, Qi Qiu, Kai Yang, Yifan Luo, Chong Chen, Tianyi Wagle, Justin Franklin, Tim Guo, Baining |
| contents | With the development of multimodal reasoning models, Computer Use Agents (CUAs), akin to Jarvis from \textit{"Iron Man"}, are becoming a reality. GUI grounding is a core component for CUAs to execute actual actions, similar to mechanical control in robotics, and it directly leads to the success or failure of the system. It determines actions such as clicking and typing, as well as related parameters like the coordinates for clicks. Current end-to-end grounding models still achieve less than 65\% accuracy on challenging benchmarks like ScreenSpot-pro and UI-Vision, indicating they are far from being ready for deployment. % , as a single misclick can result in unacceptable consequences. In this work, we conduct an empirical study on the training of grounding models, examining details from data collection to model training. Ultimately, we developed the \textbf{Phi-Ground} model family, which achieves state-of-the-art performance across all five grounding benchmarks for models under $10B$ parameters in agent settings. In the end-to-end model setting, our model still achieves SOTA results with scores of \textit{\textbf{43.2}} on ScreenSpot-pro and \textit{\textbf{27.2}} on UI-Vision. We believe that the various details discussed in this paper, along with our successes and failures, not only clarify the construction of grounding models but also benefit other perception tasks. Project homepage: \href{https://zhangmiaosen2000.github.io/Phi-Ground/}{https://zhangmiaosen2000.github.io/Phi-Ground/} |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_23779 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Phi-Ground Tech Report: Advancing Perception in GUI Grounding Zhang, Miaosen Xu, Ziqiang Zhu, Jialiang Dai, Qi Qiu, Kai Yang, Yifan Luo, Chong Chen, Tianyi Wagle, Justin Franklin, Tim Guo, Baining Computer Vision and Pattern Recognition Artificial Intelligence Multimedia With the development of multimodal reasoning models, Computer Use Agents (CUAs), akin to Jarvis from \textit{"Iron Man"}, are becoming a reality. GUI grounding is a core component for CUAs to execute actual actions, similar to mechanical control in robotics, and it directly leads to the success or failure of the system. It determines actions such as clicking and typing, as well as related parameters like the coordinates for clicks. Current end-to-end grounding models still achieve less than 65\% accuracy on challenging benchmarks like ScreenSpot-pro and UI-Vision, indicating they are far from being ready for deployment. % , as a single misclick can result in unacceptable consequences. In this work, we conduct an empirical study on the training of grounding models, examining details from data collection to model training. Ultimately, we developed the \textbf{Phi-Ground} model family, which achieves state-of-the-art performance across all five grounding benchmarks for models under $10B$ parameters in agent settings. In the end-to-end model setting, our model still achieves SOTA results with scores of \textit{\textbf{43.2}} on ScreenSpot-pro and \textit{\textbf{27.2}} on UI-Vision. We believe that the various details discussed in this paper, along with our successes and failures, not only clarify the construction of grounding models but also benefit other perception tasks. Project homepage: \href{https://zhangmiaosen2000.github.io/Phi-Ground/}{https://zhangmiaosen2000.github.io/Phi-Ground/} |
| title | Phi-Ground Tech Report: Advancing Perception in GUI Grounding |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Multimedia |
| url | https://arxiv.org/abs/2507.23779 |