{"id":13415349,"url":"https://github.com/youngwanLEE/centermask2","last_synced_at":"2025-03-14T22:33:19.859Z","repository":{"id":37625839,"uuid":"241833366","full_name":"youngwanLEE/centermask2","owner":"youngwanLEE","description":"[CVPR 2020] CenterMask : Real-time Anchor-Free Instance Segmentation","archived":false,"fork":false,"pushed_at":"2021-12-27T06:21:35.000Z","size":112,"stargazers_count":774,"open_issues_count":34,"forks_count":159,"subscribers_count":16,"default_branch":"master","last_synced_at":"2024-07-31T21:53:42.717Z","etag":null,"topics":["anchor-free","centermask","cvpr2020","detectron2","instance-segmentation","object-detection","pytorch","real-time","vovnet","vovnetv2"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/youngwanLEE.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}},"created_at":"2020-02-20T08:39:46.000Z","updated_at":"2024-07-27T10:11:15.000Z","dependencies_parsed_at":"2022-08-24T14:25:32.526Z","dependency_job_id":null,"html_url":"https://github.com/youngwanLEE/centermask2","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/youngwanLEE%2Fcentermask2","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/youngwanLEE%2Fcentermask2/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/youngwanLEE%2Fcentermask2/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/youngwanLEE%2Fcentermask2/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/youngwanLEE","download_url":"https://codeload.github.com/youngwanLEE/centermask2/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":243658057,"owners_count":20326459,"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":["anchor-free","centermask","cvpr2020","detectron2","instance-segmentation","object-detection","pytorch","real-time","vovnet","vovnetv2"],"created_at":"2024-07-30T21:00:47.452Z","updated_at":"2025-03-14T22:33:19.852Z","avatar_url":"https://github.com/youngwanLEE.png","language":"Python","funding_links":[],"categories":["Frameworks","Python"],"sub_categories":[],"readme":"# [CenterMask](https://arxiv.org/abs/1911.06667)2\n\n[[`CenterMask(original code)`](https://github.com/youngwanLEE/CenterMask)][[`vovnet-detectron2`](https://github.com/youngwanLEE/vovnet-detectron2)][[`arxiv`](https://arxiv.org/abs/1911.06667)] [[`BibTeX`](#CitingCenterMask)]\n\n**CenterMask2** is an upgraded implementation on top of [detectron2](https://github.com/facebookresearch/detectron2) beyond original [CenterMask](https://github.com/youngwanLEE/CenterMask) based on [maskrcnn-benchmark](https://github.com/facebookresearch/maskrcnn-benchmark).\n\n\u003e **[CenterMask : Real-Time Anchor-Free Instance Segmentation](https://arxiv.org/abs/1911.06667) (CVPR 2020)**\u003cbr\u003e\n\u003e [Youngwan Lee](https://github.com/youngwanLEE) and Jongyoul Park\u003cbr\u003e\n\u003e Electronics and Telecommunications Research Institute (ETRI)\u003cbr\u003e\n\u003e pre-print : https://arxiv.org/abs/1911.06667\n\n\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"https://dl.dropbox.com/s/yg9zr1tvljoeuyi/architecture.png\" width=\"850px\" /\u003e\n\u003c/div\u003e\n\n  \n  \n\n## Highlights\n- ***First* anchor-free one-stage instance segmentation.** To the best of our knowledge, **CenterMask** is the first instance segmentation on top of anchor-free object detection (15/11/2019).\n- **Toward Real-Time: CenterMask-Lite.**  This works provide not only large-scale CenterMask but also lightweight CenterMask-Lite that can run at real-time speed (\u003e 30 fps).\n- **State-of-the-art performance.**  CenterMask outperforms Mask R-CNN, TensorMask, and ShapeMask at much faster speed and CenterMask-Lite models also surpass YOLACT or YOLACT++ by large margins.\n- **Well balanced (speed/accuracy) backbone network, VoVNetV2.**  VoVNetV2 shows better performance and faster speed than ResNe(X)t or HRNet.\n\n\n## Updates\n- CenterMask2 has been released. (20/02/2020)\n- Lightweight VoVNet has ben released. (26/02/2020)\n- Panoptic-CenterMask has been released. (31/03/2020)\n- code update for compatibility with pytorch1.7 and the latest detectron2 (22/12/2020)\n## Results on COCO val\n\n### Note\n\nWe measure the inference time of all models with batch size 1 on the same V100 GPU machine.\n\n- pytorch1.7.0\n- CUDA 10.1\n- cuDNN 7.3\n- multi-scale augmentation\n- Unless speficified, no Test-Time Augmentation (TTA)\n\n\n\n### CenterMask\n\n|Method|Backbone|lr sched|inference time|mask AP|box AP|download|\n|:--------:|:--------:|:--:|:--:|----|----|:--------:|\nMask R-CNN (detectron2)|R-50|3x|0.055|37.2|41.0|\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x/137849600/model_final_f10217.pkl\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x/137849600/metrics.json\"\u003emetrics\u003c/a\u003e\nMask R-CNN (detectron2)|V2-39|3x|0.052|39.3|43.8|\u003ca href=\"https://dl.dropbox.com/s/dkto39ececze6l4/faster_V_39_eSE_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/dx9qz1dn65ccrwd/faster_V_39_eSE_ms_3x_metrics.json\"\u003emetrics\u003c/a\u003e\nCenterMask (maskrcnn-benchmark)|V2-39|3x|0.070|38.5|43.5|[link](https://github.com/youngwanLEE/CenterMask#coco-val2017-results)\n**CenterMask2**|V2-39|3x|**0.050**|**39.7**|**44.2**|\u003ca href=\"https://dl.dropbox.com/s/tczecsdxt10uai5/centermask2-V-39-eSE-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/rhoo6vkvh7rjdf9/centermask2-V-39-eSE-FPN-ms-3x_metrics.json\"\u003emetrics\u003c/a\u003e\n||\nMask R-CNN (detectron2)|R-101|3x|0.070|38.6|42.9|\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-InstanceSegmentation/mask_rcnn_R_101_FPN_3x/138205316/model_final_a3ec72.pkl\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-InstanceSegmentation/mask_rcnn_R_101_FPN_3x/138205316/metrics.json\"\u003emetrics\u003c/a\u003e\nMask R-CNN (detectron2)|V2-57|3x|0.058|39.7|44.2|\u003ca href=\"https://dl.dropbox.com/s/c7mb1mq10eo4pzk/faster_V_57_eSE_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/3tsn218zzmuhyo8/faster_V_57_eSE_metrics.json\"\u003emetrics\u003c/a\u003e\nCenterMask (maskrcnn-benchmark)|V2-57|3x|0.076|39.4|44.6|[link](https://github.com/youngwanLEE/CenterMask#coco-val2017-results)\n**CenterMask2**|V2-57|3x|**0.058**|**40.5**|**45.1**|\u003ca href=\"https://dl.dropbox.com/s/lw8nxajv1tim8gr/centermask2-V-57-eSE-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/x7r5ys3c81ldgq0/centermask2-V-57-eSE-FPN-ms-3x_metrics.json\"\u003emetrics\u003c/a\u003e\n||\nMask R-CNN (detectron2)|X-101|3x|0.129|39.5|44.3|\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-InstanceSegmentation/mask_rcnn_X_101_32x8d_FPN_3x/139653917/model_final_2d9806.pkl\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-InstanceSegmentation/mask_rcnn_X_101_32x8d_FPN_3x/139653917/metrics.json\"\u003emetrics\u003c/a\u003e\nMask R-CNN (detectron2)|V2-99|3x|0.076|40.3|44.9|\u003ca href=\"https://dl.dropbox.com/s/v64mknwzfpmfcdh/faster_V_99_eSE_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/zvaz9s8gvq2mhrd/faster_V_99_eSE_ms_3x_metrics.json\"\u003emetrics\u003c/a\u003e\nCenterMask (maskrcnn-benchmark)|V2-99|3x|0.106|40.2|45.6|[link](https://github.com/youngwanLEE/CenterMask#coco-val2017-results)\n**CenterMask2**|V2-99|3x|**0.077**|**41.4**|**46.0**|\u003ca href=\"https://dl.dropbox.com/s/c6n79x83xkdowqc/centermask2-V-99-eSE-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/jdzgmdatit00hq5/centermask2-V-99-eSE-FPN-ms-3x_metrics.json\"\u003emetrics\u003c/a\u003e\n||\n**CenterMask2 (TTA)**|V2-99|3x|-|**42.5**|**48.6**|\u003ca href=\"https://dl.dropbox.com/s/c6n79x83xkdowqc/centermask2-V-99-eSE-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/jdzgmdatit00hq5/centermask2-V-99-eSE-FPN-ms-3x_metrics.json\"\u003emetrics\u003c/a\u003e\n* TTA denotes Test-Time Augmentation (multi-scale test).\n\n### CenterMask-Lite\n\n|Method|Backbone|lr sched|inference time|mask AP|box AP|download|\n|:--------:|:--------:|:--:|:--:|:----:|:----:|:--------:|\n|YOLACT550|R-50|4x|0.023|28.2|30.3|[link](https://github.com/dbolya/yolact)\n|CenterMask (maskrcnn-benchmark)|V-19|4x|0.023|32.4|35.9|[link](https://github.com/youngwanLEE/CenterMask#coco-val2017-results)\n|**CenterMask2-Lite**|V-19|4x|0.023|**32.8**|**35.9**|\u003ca href=\"https://dl.dropbox.com/s/dret2ap7djty7mp/centermask2-lite-V-19-eSE-FPN-ms-4x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/zsta7azy87a833u/centermask2-lite-V-19-eSE-FPN-ms-4x-metrics.json\"\u003emetrics\u003c/a\u003e\n||\n|YOLACT550|R-101|4x|0.030|28.2|30.3|[link](https://github.com/dbolya/yolact)\n|YOLACT550++|R-50|4x|0.029|34.1|-|[link](https://github.com/dbolya/yolact)\n|YOLACT550++|R-101|4x|0.036|34.6|-|[link](https://github.com/dbolya/yolact)\n|CenterMask (maskrcnn-benchmark)|V-39|4x|0.027|36.3|40.7|[link](https://github.com/youngwanLEE/CenterMask#coco-val2017-results)\n|**CenterMask2-Lite**|V-39|4x|0.028|**36.7**|**40.9**|\u003ca href=\"https://dl.dropbox.com/s/uwc0ypa1jvco2bi/centermask2-lite-V-39-eSE-FPN-ms-4x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/aoa6y3i3el4edbk/centermask2-lite-V-39-eSE-FPN-ms-4x-metrics.json\"\u003emetrics\u003c/a\u003e\n* Note that The inference time is measured on Titan Xp GPU for fair comparison with YOLACT.\n\n### Lightweight VoVNet backbone\n\n|Method|Backbone|Param.|lr sched|inference time|mask AP|box AP|download|\n|:--------:|:--------:|:--:|:--:|:--:|:----:|:----:|:--------:|\n|CenterMask2-Lite|MobileNetV2|3.5M|4x|0.021|27.2|29.8|\u003ca href=\"https://dl.dropbox.com/s/8omou546f0n78nj/centermask_lite_Mv2_ms_4x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/2jlcwy30eq72w47/centermask_lite_Mv2_ms_4x_metrics.json\"\u003emetrics\u003c/a\u003e\n||\n|CenterMask2-Lite|V-19|11.2M|4x|0.023|32.8|35.9|\u003ca href=\"https://dl.dropbox.com/s/dret2ap7djty7mp/centermask2-lite-V-19-eSE-FPN-ms-4x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/zsta7azy87a833u/centermask2-lite-V-19-eSE-FPN-ms-4x-metrics.json\"\u003emetrics\u003c/a\u003e\n|CenterMask2-Lite|V-19-**Slim**|3.1M|4x|0.021|29.8|32.5|\u003ca href=\"https://dl.dropbox.com/s/o2n1ifl0zkbv16x/centermask-lite-V-19-eSE-slim-FPN-ms-4x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/8y71oz0kxwqk7go/centermask-lite-V-19-eSE-slim-FPN-ms-4x-metrics.json?dl=0\"\u003emetrics\u003c/a\u003e\n|CenterMask2-Lite|V-19**Slim**-**DW**|1.8M|4x|0.020|27.1|29.5|\u003ca href=\"https://dl.dropbox.com/s/vsvhwtqm6ko1c7m/centermask-lite-V-19-eSE-slim-dw-FPN-ms-4x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/q4idjnsgvo151zx/centermask-lite-V-19-eSE-slim-dw-FPN-ms-4x-metrics.json\"\u003emetrics\u003c/a\u003e\n* _**DW** and **Slim** denote depthwise separable convolution and a thiner model with half the channel size, respectively._   \n* __Params.__ means the number of parameters of backbone.   \n\n### Deformable VoVNet Backbone\n\n|Method|Backbone|lr sched|inference time|mask AP|box AP|download|\n|:--------:|:--------:|:--:|:--:|:--:|:----:|:----:|\nCenterMask2|V2-39|3x|0.050|39.7|44.2|\u003ca href=\"https://dl.dropbox.com/s/tczecsdxt10uai5/centermask2-V-39-eSE-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/rhoo6vkvh7rjdf9/centermask2-V-39-eSE-FPN-ms-3x_metrics.json\"\u003emetrics\u003c/a\u003e\nCenterMask2|V2-39-DCN|3x|0.061|40.3|45.1|\u003ca href=\"https://dl.dropbox.com/s/zmps03vghzirk7v/centermask-V-39-eSE-dcn-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/aj1mr8m32z11zbw/centermask-V-39-eSE-dcn-FPN-ms-3x-metrics.json\"\u003emetrics\u003c/a\u003e\n||\nCenterMask2|V2-57|3x|0.058|40.5|45.1|\u003ca href=\"https://dl.dropbox.com/s/lw8nxajv1tim8gr/centermask2-V-57-eSE-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/x7r5ys3c81ldgq0/centermask2-V-57-eSE-FPN-ms-3x_metrics.json\"\u003emetrics\u003c/a\u003e\nCenterMask2|V2-57-DCN|3x|0.071|40.9|45.5|\u003ca href=\"https://dl.dropbox.com/s/1f64azqyd2ot6qq/centermask-V-57-eSE-dcn-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/b3zpguko137r6eh/centermask-V-57-eSE-dcn-FPN-ms-3x-metrics.json\"\u003emetrics\u003c/a\u003e\n||\nCenterMask2|V2-99|3x|0.077|41.4|46.0|\u003ca href=\"https://dl.dropbox.com/s/c6n79x83xkdowqc/centermask2-V-99-eSE-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/jdzgmdatit00hq5/centermask2-V-99-eSE-FPN-ms-3x_metrics.json\"\u003emetrics\u003c/a\u003e\nCenterMask2|V2-99-DCN|3x|0.110|42.0|46.9|\u003ca href=\"https://dl.dropbox.com/s/atuph90nzm7s8x8/centermask-V-99-eSE-dcn-FPN-ms-3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/82ulexlivy19cve/centermask-V-99-eSE-dcn-FPN-ms-3x-metrics.json\"\u003emetrics\u003c/a\u003e\n||\n\n* _DCN denotes deformable convolutional networks v2. Note that we apply deformable convolutions from stage 3 to 5 in backbones._\n\n### Panoptic-CenterMask\n\n|Method|Backbone|lr sched|inference time|mask AP|box AP|PQ|download|\n|:--------:|:--------:|:--:|:--:|:--:|:----:|:----:|:--------:|\n|Panoptic-FPN|R-50|3x|0.063|40.0|36.5|41.5|\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-PanopticSegmentation/panoptic_fpn_R_50_3x/139514569/model_final_c10459.pkl\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-PanopticSegmentation/panoptic_fpn_R_50_3x/139514569/metrics.json\"\u003emetrics\u003c/a\u003e\n|Panoptic-CenterMask|R-50|3x|0.063|41.4|37.3|42.0|\u003ca href=\"https://dl.dropbox.com/s/vxe51cdeprao94j/panoptic_centermask_R_50_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/dfddgx6rnw1zr4l/panoptic_centermask_R_50_ms_3x_metrics.json\"\u003emetrics\u003c/a\u003e\n|Panoptic-FPN|V-39|3x|0.063|42.8|38.5|43.4|\u003ca href=\"https://dl.dropbox.com/s/fnr9r4arv0cbfbf/panoptic_V_39_eSE_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/vftfukrjuu7w1ao/panoptic_V_39_eSE_3x_metrics.json\"\u003emetrics\u003c/a\u003e\n|Panoptic-CenterMask|V-39|3x|0.066|43.4|39.0|43.7|\u003ca href=\"https://dl.dropbox.com/s/49ig16ailra1f4t/panoptic_centermask_V_39_eSE_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/wy4mn8n513k0um5/panoptic_centermask_V_39_eSE_ms_3x_metrics.json\"\u003emetrics\u003c/a\u003e\n||\n|Panoptic-FPN|R-101|3x|0.078|42.4|38.5|43.0|\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-PanopticSegmentation/panoptic_fpn_R_101_3x/139514519/model_final_cafdb1.pkl\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.fbaipublicfiles.com/detectron2/COCO-PanopticSegmentation/panoptic_fpn_R_101_3x/139514519/metrics.json\"\u003emetrics\u003c/a\u003e\n|Panoptic-CenterMask|R-101|3x|0.076|43.5|39.0|43.6|\u003ca href=\"https://dl.dropbox.com/s/y5stg3qx72gff5o/panoptic_centermask_R_101_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/ojljt0obp8vnr8s/panoptic_centermask_R_101_ms_3x_metrics.json\"\u003emetrics\u003c/a\u003e\n|Panoptic-FPN|V-57|3x|0.070|43.4|39.2|44.3|\u003ca href=\"https://www.dropbox.com/s/zhoqx5rvc0jj0oa/panoptic_V_57_eSE_3x.pth?dl=1\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/20hwrmru15dilre/panoptic_V_57_eSE_3x_metrics.json\"\u003emetrics\u003c/a\u003e\n|Panoptic-CenterMask|V-57|3x|0.071|43.9|39.6|44.5|\u003ca href=\"https://dl.dropbox.com/s/kqukww4y7tbgbrh/panoptic_centermask_V_57_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/4asto3b4iya74ak/panoptic_centermask_V_57_ms_3x_metrics.json\"\u003emetrics\u003c/a\u003e\n||\n|Panoptic-CenterMask|V-99|3x|0.091|45.1|40.6|45.4|\u003ca href=\"https://dl.dropbox.com/s/pr6a3inpasn7qlz/panoptic_centermask_V_99_ms_3x.pth\"\u003emodel\u003c/a\u003e\u0026nbsp;\\|\u0026nbsp;\u003ca href=\"https://dl.dropbox.com/s/00e8x0riplme7pm/panoptic_centermask_V_99_ms_3x_metrics.json\"\u003emetrics\u003c/a\u003e\n\n\n## Installation\nAll you need to use centermask2 is [detectron2](https://github.com/facebookresearch/detectron2). It's easy!    \nyou just install [detectron2](https://github.com/facebookresearch/detectron2) following [INSTALL.md](https://github.com/facebookresearch/detectron2/blob/master/INSTALL.md).   \nPrepare for coco dataset following [this instruction](https://github.com/facebookresearch/detectron2/tree/master/datasets).\n\n## Training\n\n#### ImageNet Pretrained Models\n\nWe provide backbone weights pretrained on ImageNet-1k dataset for detectron2.\n* [MobileNet-V2](https://dl.dropbox.com/s/yduxbc13s3ip6qn/mobilenet_v2_detectron2.pth)\n* [VoVNetV2-19-Slim-DW](https://dl.dropbox.com/s/f3s7ospitqoals1/vovnet19_ese_slim_dw_detectron2.pth)\n* [VoVNetV2-19-Slim](https://dl.dropbox.com/s/8h5ybmi4ftbcom0/vovnet19_ese_slim_detectron2.pth)\n* [VoVNetV2-19](https://dl.dropbox.com/s/rptgw6stppbiw1u/vovnet19_ese_detectron2.pth)\n* [VoVNetV2-39](https://dl.dropbox.com/s/q98pypf96rhtd8y/vovnet39_ese_detectron2.pth)\n* [VoVNetV2-57](https://dl.dropbox.com/s/8xl0cb3jj51f45a/vovnet57_ese_detectron2.pth)\n* [VoVNetV2-99](https://dl.dropbox.com/s/1mlv31coewx8trd/vovnet99_ese_detectron2.pth)\n\n\nTo train a model, run\n```bash\ncd centermask2\npython train_net.py --config-file \"configs/\u003cconfig.yaml\u003e\"\n```\n\nFor example, to launch CenterMask training with VoVNetV2-39 backbone on 8 GPUs,\none should execute:\n```bash\ncd centermask2\npython train_net.py --config-file \"configs/centermask/centermask_V_39_eSE_FPN_ms_3x.yaml\" --num-gpus 8\n```\n\n## Evaluation\n\nModel evaluation can be done similarly:   \n* if you want to inference with 1 batch `--num-gpus 1` \n* `--eval-only`\n* `MODEL.WEIGHTS path/to/the/model.pth`\n\n```bash\ncd centermask2\nwget https://dl.dropbox.com/s/tczecsdxt10uai5/centermask2-V-39-eSE-FPN-ms-3x.pth\npython train_net.py --config-file \"configs/centermask/centermask_V_39_eSE_FPN_ms_3x.yaml\" --num-gpus 1 --eval-only MODEL.WEIGHTS centermask2-V-39-eSE-FPN-ms-3x.pth\n```\n\n## TODO\n - [x] Adding Lightweight models\n - [ ] Applying CenterMask for PointRend or Panoptic-FPN.\n\n\n## \u003ca name=\"CitingCenterMask\"\u003e\u003c/a\u003eCiting CenterMask\n\nIf you use VoVNet, please use the following BibTeX entry.\n\n```BibTeX\n@inproceedings{lee2019energy,\n  title = {An Energy and GPU-Computation Efficient Backbone Network for Real-Time Object Detection},\n  author = {Lee, Youngwan and Hwang, Joong-won and Lee, Sangrok and Bae, Yuseok and Park, Jongyoul},\n  booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops},\n  year = {2019}\n}\n\n@inproceedings{lee2020centermask,\n  title={CenterMask: Real-Time Anchor-Free Instance Segmentation},\n  author={Lee, Youngwan and Park, Jongyoul},\n  booktitle={CVPR},\n  year={2020}\n}\n```\n\n## Special Thanks to\n\n[mask scoring for detectron2](https://github.com/lsrock1/maskscoring_rcnn.detectron2) by [Sangrok Lee](https://github.com/lsrock1)   \n[FCOS_for_detectron2](https://github.com/aim-uofa/adet) by AdeliDet team.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FyoungwanLEE%2Fcentermask2","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FyoungwanLEE%2Fcentermask2","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FyoungwanLEE%2Fcentermask2/lists"}