{"id":28420022,"url":"https://github.com/aim-uofa/active-o3","last_synced_at":"2025-06-28T19:32:13.183Z","repository":{"id":295915858,"uuid":"989947184","full_name":"aim-uofa/Active-o3","owner":"aim-uofa","description":"ACTIVE-O3: Empowering Multimodal Large Language Models with Active Perception via GRPO","archived":false,"fork":false,"pushed_at":"2025-05-30T02:48:15.000Z","size":5093,"stargazers_count":58,"open_issues_count":2,"forks_count":1,"subscribers_count":2,"default_branch":"main","last_synced_at":"2025-06-05T03:14:57.144Z","etag":null,"topics":["active-perception","active-vision","grpo","mllms","o3","rl","thinking-with-image"],"latest_commit_sha":null,"homepage":"https://aim-uofa.github.io/ACTIVE-o3/","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/aim-uofa.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-05-25T07:10:15.000Z","updated_at":"2025-06-03T08:00:23.000Z","dependencies_parsed_at":"2025-05-28T03:18:22.586Z","dependency_job_id":"a31017e0-492c-43ce-afa3-32cccbe3438e","html_url":"https://github.com/aim-uofa/Active-o3","commit_stats":null,"previous_names":["aim-uofa/active-o3"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/aim-uofa/Active-o3","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aim-uofa%2FActive-o3","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aim-uofa%2FActive-o3/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aim-uofa%2FActive-o3/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aim-uofa%2FActive-o3/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/aim-uofa","download_url":"https://codeload.github.com/aim-uofa/Active-o3/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aim-uofa%2FActive-o3/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":262484111,"owners_count":23318394,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["active-perception","active-vision","grpo","mllms","o3","rl","thinking-with-image"],"created_at":"2025-06-04T23:40:32.298Z","updated_at":"2025-06-28T19:32:13.173Z","avatar_url":"https://github.com/aim-uofa.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cdiv align=\"center\"\u003e\n\n# ACTIVE-O3: Empowering Multimodal Large Language Models with Active Perception via GRPO\n\n\n\n\u003csup\u003e1\u003c/sup\u003e[Zhejiang University](https://www.zju.edu.cn/english/), \u0026nbsp;\n\u003csup\u003e2\u003c/sup\u003e[Ant Group](https://www.antgroup.com/en)\n\n\u003c!-- **CVPR2025** --\u003e\n\n[📄 **Paper**](https://arxiv.org/abs/2505.21457)\u0026nbsp; | \u0026nbsp;[🌐 **Project Page**](https://aim-uofa.github.io/ACTIVE-o3)\u0026nbsp; | \u0026nbsp;[💾 **Model Weights**](https://www.modelscope.cn/models/zzzmmz/ACTIVE-o3)\n\u003c/div\u003e\n\n## 🚀 Overview\n\u003cdiv align=\"center\"\u003e\n\u003cimg width=\"800\" alt=\"SegAgent Framework\" src=\"images/frame.jpeg\"\u003e\n\u003c/div\u003e\n\n## 📖 Description\n\nwe propose ACTIVE-O3, a purely reinforcement learning-based training framework built on top of GRPO, designed to equip MLLMs with active perception capabilities. We further establish a comprehensive benchmark suite to evaluate ACTIVE-O3 across both general open-world tasks—such as small-object and dense object grounding—and domain-specific scenarios, including small object detection in remote sensing and autonomous driving, as well as fine-grained interactive segmentation. Experimental results demonstrate that ACTIVE-O3 significantly enhances active perception capabilities compared to Qwen-VL2.5-CoT. For example, Figure 1 shows an example of zero-shot reasoning on the V* benchmark, where ACTIVE- O3 successfully identifies the number on the traffic light by zooming in on the relevant region, while Qwen2.5-VL fails to do so. Moreover, across all downstream tasks, ACTIVE-O3 consistently improves performance under fixed computational budgets. We hope that our work here can provide a simple codebase and evaluation protocol to facilitate future research on active perception MLLM.\n\n\n\n\n## 🚩 Plan\n\u003c!-- - [ ] Release the weights. --\u003e\n- [x] Release the weights.\n- [x] Release the inference demo.\n- [ ] Release the dataset.\n- [ ] Release the training scripts.\n- [ ] Release the evaluation scripts.\n\u003c!-- --- --\u003e\n\n\n\n## 🛠️ Getting Started\n\n### 📐 Set up Environment\n\n```bash\n\n# build environment\nconda create -n activeo3 python=3.10\nconda activate activeo3\n\n# install packages\npip install torch==2.5.1 torchvision==0.20.1\npip install flash-attn --no-build-isolation\npip install transformers==4.51.3\npip install qwen-omni-utils[decord]\n```\n###  🔍 demo\n\n```bash\n# run demo\npython demo/activeo3_demo_vstar.py\n```\n\n\n## 🎫 License\n\nFor academic usage, this project is licensed under [the 2-clause BSD License](LICENSE). For commercial inquiries, please contact [Chunhua Shen](mailto:chhshen@gmail.com).\n\n## 🖊️ Citation\n\nIf you find this work helpful for your research, please cite:\n\n```BibTeX\n@article{zhu2025active,\n  title={Active-O3: Empowering Multimodal Large Language Models with Active Perception via GRPO},\n  author={Zhu, Muzhi and Zhong, Hao and Zhao, Canyu and Du, Zongze and Huang, Zheng and Liu, Mingyu and Chen, Hao and Zou, Cheng and Chen, Jingdong and Yang, Ming and others},\n  journal={arXiv preprint arXiv:2505.21457},\n  year={2025}\n}","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faim-uofa%2Factive-o3","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faim-uofa%2Factive-o3","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faim-uofa%2Factive-o3/lists"}