{"id":20962234,"url":"https://github.com/lartpang/comptr","last_synced_at":"2025-12-25T17:59:14.894Z","repository":{"id":205261672,"uuid":"669833011","full_name":"lartpang/ComPtr","owner":"lartpang","description":"ComPtr: Towards Diverse Bi-source Dense Prediction Tasks via A Simple yet General Complementary Transformer","archived":false,"fork":false,"pushed_at":"2024-02-08T08:21:47.000Z","size":5,"stargazers_count":7,"open_issues_count":2,"forks_count":1,"subscribers_count":2,"default_branch":"main","last_synced_at":"2025-01-20T00:53:44.175Z","etag":null,"topics":["change-detection","change-detection-algorithms","change-detector","crowd-count","crowd-counting","crowdcounter","crowdcounting","rgbd-saliency-detection","rgbd-salient-object-detection","rgbd-segmentation","rgbd-semantic-segmentation","rgbt-saliency-detection","rgbt-salient-object-detection","saliency-detection","saliency-methods","saliency-model","salient-area-extraction","salient-object-detection"],"latest_commit_sha":null,"homepage":"https://arxiv.org/abs/2307.12349","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/lartpang.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,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2023-07-23T15:20:01.000Z","updated_at":"2024-02-26T03:39:57.000Z","dependencies_parsed_at":"2024-12-06T17:40:18.648Z","dependency_job_id":"588f3002-97fb-4224-961e-fcc7fb0aa983","html_url":"https://github.com/lartpang/ComPtr","commit_stats":{"total_commits":3,"total_committers":1,"mean_commits":3.0,"dds":0.0,"last_synced_commit":"cb2165913199ab8ac86ae58106e59b860f97490b"},"previous_names":["lartpang/comptr"],"tags_count":1,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lartpang%2FComPtr","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lartpang%2FComPtr/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lartpang%2FComPtr/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lartpang%2FComPtr/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/lartpang","download_url":"https://codeload.github.com/lartpang/ComPtr/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":243358307,"owners_count":20277995,"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":["change-detection","change-detection-algorithms","change-detector","crowd-count","crowd-counting","crowdcounter","crowdcounting","rgbd-saliency-detection","rgbd-salient-object-detection","rgbd-segmentation","rgbd-semantic-segmentation","rgbt-saliency-detection","rgbt-salient-object-detection","saliency-detection","saliency-methods","saliency-model","salient-area-extraction","salient-object-detection"],"created_at":"2024-11-19T02:25:09.653Z","updated_at":"2025-12-25T17:59:14.889Z","avatar_url":"https://github.com/lartpang.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"# ComPtr\n\nComPtr: Towards Diverse Bi-source Dense Prediction Tasks via A Simple yet General Complementary Transformer\n\n```bibtex\n@ARTICLE{ComPtr,\n    title={ComPtr: Towards Diverse Bi-source Dense Prediction Tasks via A Simple yet General Complementary Transformer},\n    author={Youwei Pang and Xiaoqi Zhao and Lihe Zhang and Huchuan Lu},\n    journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},\n    year    = {2025},\n    doi     = {10.1109/TPAMI.2025.3578494},\n}\n```\n\n## Results\n\n**LEVIR-CD**\n\n\u003e Based on [CAVER](https://github.com/lartpang/CAVER).\n\n| Methods  | Backbone       | F1 $\\uparrow$ | IoU $\\uparrow$ |\n| -------- | -------------- | ------------- | -------------- |\n| TTP      | SAM (ViT-L/16) | 92.10         | 85.60          |\n| ComPtr-T | Swin-T         | 92.62         | 86.25          |\n| ComPtr-B | Swin-B         | 92.70         | 86.39          |\n\n**SYSU-CD**\n\n\u003e Based on [CAVER](https://github.com/lartpang/CAVER).\n\n| Methods    | Backbone | F1 $\\uparrow$ | IoU $\\uparrow$ |\n| ---------- | -------- | ------------- | -------------- |\n| TransY-Net | Swin-B   | 82.84         | 70.71          |\n| ComPtr-T   | Swin-T   | 83.43         | 71.57          |\n| ComPtr-B   | Swin-B   | 83.18         | 71.21          |\n\n**RGB-T CC**\n\n\u003e Based on [MMCC](https://github.com/liuzywen/RGBTCC).\n\n| Methods  | Backbone | GAME $_{0}\\downarrow$ | RMSE $\\downarrow$ |\n| -------- | -------- | --------------------- | ----------------- |\n| MMCC     | PvTv2-B3 | 10.90                 | 18.79             |\n| ComPtr-T | Swin-T   | 10.52                 | 18.48             |\n| ComPtr-B | Swin-B   | 11.82                 | 20.75             |\n\n**SUN-RGBD**\n\n\u003e Based on [CMX](https://github.com/huaaaliu/RGBX_Semantic_Segmentation).\n\n| Methods   | Backbone  | mIoU $\\uparrow$ |\n| --------- | --------- | --------------- |\n| DFormer-L | DFormer-L | 52.5            |\n| ComPtr-T  | Swin-T    | 48.9            |\n| ComPtr-B  | Swin-B    | 52.8            |\n\n**RGB-D/T SOD**\n\n\u003e Based on [CAVER](https://github.com/lartpang/CAVER).\n\nSee [Releases](https://github.com/lartpang/ComPtr/releases/tag/assets-v1.0)\n\n(S $\\_m$ \u0026 F $^{\\omega}\\_{\\beta}$)\n\n| Methods    | Backbone   | DUTLF-Depth $\\uparrow$ | NJUD $\\uparrow$ | NLPR $\\uparrow$ | SIP $\\uparrow$ | STEREO1000 $\\uparrow$ |\n| ---------- | ---------- | ---------------------- | --------------- | --------------- | -------------- | --------------------- |\n| HRTransNet | HRFormer-B | 95.07 \u0026 94.93          | 93.33 \u0026 92.00   | 94.18 \u0026 91.70   | 90.90 \u0026 89.56  | 92.12 \u0026 89.37         |  |\n| ComPtr-T   | Swin-T     | 94.37 \u0026 92.17          | 92.89 \u0026 89.48   | 93.56 \u0026 88.66   | 91.34 \u0026 87.96  | 91.99 \u0026 87.45         |\n| ComPtr-B   | Swin-B     | 95.87 \u0026 95.14          | 93.96 \u0026 91.95   | 94.29 \u0026 90.73   | 91.53 \u0026 89.48  | 93.29 \u0026 90.29         |\n\n| Methods    | Backbone   | VT821 $\\uparrow$ | VT1000 $\\uparrow$ | VT5000 $\\uparrow$ |\n| ---------- | ---------- | ---------------- | ----------------- | ----------------- |\n| HRTransNet | HRFormer-B | 90.59 \u0026 84.92    | 93.79 \u0026 91.27     | 91.23 \u0026 86.99     |\n| ComPtr-T   | Swin-T     | 90.52 \u0026 83.32    | 94.17 \u0026 89.82     | 90.71 \u0026 83.10     |\n| ComPtr-B   | Swin-B     | 92.37 \u0026 87.48    | 95.12 \u0026 92.66     | 92.84 \u0026 88.34     |\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flartpang%2Fcomptr","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flartpang%2Fcomptr","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flartpang%2Fcomptr/lists"}