{"id":13499095,"url":"https://github.com/MMLab-CU/polynet","last_synced_at":"2025-03-29T03:32:19.070Z","repository":{"id":79472889,"uuid":"87809975","full_name":"MMLab-CU/polynet","owner":"MMLab-CU","description":"The Github Repo for PolyNet","archived":false,"fork":false,"pushed_at":"2017-07-23T07:37:53.000Z","size":364,"stargazers_count":82,"open_issues_count":4,"forks_count":16,"subscribers_count":17,"default_branch":"master","last_synced_at":"2024-12-12T02:02:18.298Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/MMLab-CU.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null}},"created_at":"2017-04-10T12:47:35.000Z","updated_at":"2024-10-14T17:31:56.000Z","dependencies_parsed_at":"2023-07-13T19:45:43.995Z","dependency_job_id":null,"html_url":"https://github.com/MMLab-CU/polynet","commit_stats":null,"previous_names":["open-mmlab/polynet"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MMLab-CU%2Fpolynet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MMLab-CU%2Fpolynet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MMLab-CU%2Fpolynet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MMLab-CU%2Fpolynet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/MMLab-CU","download_url":"https://codeload.github.com/MMLab-CU/polynet/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246135766,"owners_count":20729056,"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":[],"created_at":"2024-07-31T22:00:28.522Z","updated_at":"2025-03-29T03:32:18.823Z","avatar_url":"https://github.com/MMLab-CU.png","language":null,"funding_links":[],"categories":["Papers\u0026Codes"],"sub_categories":["PolyNet"],"readme":"# PolyNet: A Pursuit of Structural Diversity in Very Deep Networks\n\nBy [Xingcheng Zhang](https://www.linkedin.com/in/xingchengzhang/), [Zhizhong Li](http://mmlab.ie.cuhk.edu.hk/html_people/postgraduate_Zhizhong_Li.html), [Chen Change Loy](http://personal.ie.cuhk.edu.hk/~ccloy/), [Dahua Lin](http://dahua.me)\n\n[Multimedia Laboratory](http://mmlab.ie.cuhk.edu.hk), The Chinese University of Hong Kong\n\n[![Homepage](https://img.shields.io/badge/Project-Homepage-orange.svg?style=flat\u0026colorA=E1523D\u0026colorB=007D8A)](http://mmlab.ie.cuhk.edu.hk/projects/cu_deeplink/)\n[![pdf](https://img.shields.io/badge/Arxiv-pdf-orange.svg?style=flat)](https://arxiv.org/abs/1611.05725)\n\n### BibTeX\n\n```bib\n@article{zhang2016polynet,\n  title={Polynet: A pursuit of structural diversity in very deep networks},\n  author={Zhang, Xingcheng and Li, Zhizhong and Loy, Chen Change and Lin, Dahua},\n  journal={arXiv preprint arXiv:1611.05725},\n  year={2016}\n}\n```\n\n### The Very Deep PolyNet\n  ![PolyNet](polynet.png)\n* [Visualization](http://ethereon.github.io/netscope/#/gist/b22923712859813a051c796b19ce5944)\n* Models\n    * **Parrots**: model files to be added;\n    * [Caffe](https://github.com/yjxiong/caffe): [proto](https://drive.google.com/open?id=0B6pxsvrUJ931aTJmdEJOODhtQ3c), [model](https://drive.google.com/open?id=0B6pxsvrUJ931WXluclRBVDAtaEk).\n\n  NOTE: The model is trained using our own deep learning framework **Parrots**. The caffe model is converted from the **Parrots** model, and the proto file is based on [Yuanjun's fork](https://github.com/yjxiong/caffe) of Caffe.\n\n### Results\n\n  \u003cimg src=\"https://rawgit.com/CUHK-MMLAB/polynet/master/compare.svg\" width=\"60%\"\u003e\n\nmodel|training speed* (#imgs/second)|single-crop val top-1|single-crop val top-5|single-crop test top-5 | multi-crop val top-1 | multi-crop val top-5\n:---:|:---:|:---:|:---:|:---:|:---:|:---:\n[ResNet-152](https://github.com/KaimingHe/deep-residual-networks) |-| [22.16](https://github.com/facebook/fb.resnet.torch#single-crop-224x224-validation-error-rate) | [6.16](https://github.com/facebook/fb.resnet.torch#single-crop-224x224-validation-error-rate) | - | [19.38](https://arxiv.org/pdf/1512.03385.pdf) | [4.49](https://arxiv.org/pdf/1512.03385.pdf)\nResNet-152^ |279 ( 8 GPUs)| 20.93 | 5.54 | 5.50 | 18.50 | 3.97\nResNet-269^ |245 (16 GPUs)| 19.78 | 4.89 | 4.82 | 17.54 | 3.55\nResNet-500^ |248 (32 GPUs)| 19.66 | 4.78 | 4.70 | 17.59\t| 3.63\n[Inception-v4](https://arxiv.org/abs/1602.07261) |-| 20.0 | 5.0 | -| 1 7.7 | 3.8\n[Inception-ResNet-v2](https://arxiv.org/abs/1602.07261) |-| 19.9 |  4.9 | - | 17.8 | 3.7\nInception-ResNet-v2 |314 ( 8 GPUs)| 20.05  | 5.05  | 5.11 | 18.41 | 3.98\nVery Deep Inception-ResNet |278 (32 GPUs)| 19.10 | 4.48 | 4.46 | 17.39  | 3.56\nVery Deep PolyNet |290 (32 GPUs)| **18.71** | **4.25** | **4.33** | **17.36**  | **3.45**\n\n  ^ The ResNet models are trained by [Tong Xiao](https://github.com/Cysu);\n\n  \\* Training speed is measured on **Parrots** using NVIDIA TITAN X Graphics Cards.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FMMLab-CU%2Fpolynet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FMMLab-CU%2Fpolynet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FMMLab-CU%2Fpolynet/lists"}