{"id":26159672,"url":"https://github.com/xuxw98/ESAM","last_synced_at":"2025-03-11T11:33:34.146Z","repository":{"id":254097195,"uuid":"727103787","full_name":"xuxw98/ESAM","owner":"xuxw98","description":"EmbodiedSAM: Online Segment Any 3D Thing in Real Time","archived":false,"fork":false,"pushed_at":"2024-11-01T08:19:34.000Z","size":62702,"stargazers_count":223,"open_issues_count":0,"forks_count":14,"subscribers_count":4,"default_branch":"main","last_synced_at":"2024-11-19T09:23:48.662Z","etag":null,"topics":["3d-instance-segmentation","3d-scene-understanding","embodied-vision","real-time","segment-anything","semi-supervised","streaming-video"],"latest_commit_sha":null,"homepage":"https://xuxw98.github.io/ESAM/","language":"Python","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/xuxw98.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":"2023-12-04T07:34:11.000Z","updated_at":"2024-11-19T01:53:46.000Z","dependencies_parsed_at":"2024-11-30T20:04:49.970Z","dependency_job_id":"69a51175-2672-45c6-bccc-350693f42983","html_url":"https://github.com/xuxw98/ESAM","commit_stats":null,"previous_names":["xuxw98/esam"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/xuxw98%2FESAM","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/xuxw98%2FESAM/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/xuxw98%2FESAM/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/xuxw98%2FESAM/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/xuxw98","download_url":"https://codeload.github.com/xuxw98/ESAM/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":243025961,"owners_count":20223899,"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":["3d-instance-segmentation","3d-scene-understanding","embodied-vision","real-time","segment-anything","semi-supervised","streaming-video"],"created_at":"2025-03-11T11:33:28.255Z","updated_at":"2025-03-11T11:33:34.132Z","avatar_url":"https://github.com/xuxw98.png","language":"Python","funding_links":[],"categories":["Uncategorized"],"sub_categories":["Uncategorized"],"readme":"# EmbodiedSAM: Online Segment Any 3D Thing in Real Time\r\n### [Paper](https://arxiv.org/abs/2408.11811) | [Project Page](https://xuxw98.github.io/ESAM/) | [Video](https://cloud.tsinghua.edu.cn/f/f75279f89bf64720b8ec/?dl=1) | [中文解读](https://zhuanlan.zhihu.com/p/23105869992)\r\n\r\n\u003e EmbodiedSAM: Online Segment Any 3D Thing in Real Time  \r\n\u003e [Xiuwei Xu](https://xuxw98.github.io/), Huangxing Chen, [Linqing Zhao](https://scholar.google.com/citations?user=ypxt5UEAAAAJ\u0026hl=zh-CN\u0026oi=ao), [Ziwei Wang](https://ziweiwangthu.github.io/), [Jie Zhou](https://scholar.google.com/citations?user=6a79aPwAAAAJ\u0026hl=en\u0026authuser=1), [Jiwen Lu](http://ivg.au.tsinghua.edu.cn/Jiwen_Lu/)\r\n\r\n\r\nIn this work, we presented ESAM, an efficient framework that leverages vision foundation models for \u003cb\u003eonline\u003c/b\u003e, \u003cb\u003ereal-time\u003c/b\u003e, \u003cb\u003efine-grained\u003c/b\u003e, \u003cb\u003egeneralized\u003c/b\u003e and \u003cb\u003eopen-vocabulary\u003c/b\u003e 3D instance segmentation. \r\n\r\n\r\n## News\r\n- [2025/2/11]: EmbodiedSAM is selected as an \u003cb\u003eoral presentation\u003c/b\u003e in ICLR 2025!\r\n- [2025/1/23]: EmbodiedSAM is accepted to ICLR 2025 with a top 2% rating!\r\n- [2024/10/01]: Demo code available at [here](https://github.com/xuxw98/ESAM/blob/main/docs/demo.md).\r\n- [2024/8/27]: Fix some bugs.\r\n- [2024/8/22]: Code and demo released.\r\n\r\n\r\n## Demo\r\n### Bedroom:\r\n![demo](./assets/demo2.gif)\r\n\r\n### Office:\r\n![demo](./assets/demo1.gif)\r\n\r\nDemos are a little bit large; please wait a moment to load them. Welcome to the home page for more complete demos and detailed introductions.\r\n\r\n\r\n## Method \r\n\r\nMethod Pipeline:\r\n![overview](./assets/pipeline.png)\r\n\r\n## Getting Started\r\nFor environment setup and dataset preparation, please follow:\r\n* [Installation](./docs/installation.md)\r\n* [Dataset Preparation](./docs/dataset_preparation.md)\r\n\r\nFor training and evaluation, please follow:\r\n* [Train and Evaluation](./docs/run.md)\r\n  \r\nFor visualization demo, please follow:\r\n* [Visualization Demo](./docs/demo.md)\r\n\r\n\r\n## Main Results\r\nWe provide the checkpoints for quick reproduction of the results reported in the paper.\r\n\r\n**Class-agnostic 3D instance segmentation results on ScanNet200 dataset:**\r\n\r\n|  Method  |   Type  |     VFM     |  AP  | AP@50 | AP@25 | Speed(ms) | Downloads |\r\n|:--------:|:-------:|:-----------:|:----:|:-----:|:-----:|:---------:|:---------:|\r\n| [SAMPro3D](https://github.com/GAP-LAB-CUHK-SZ/SAMPro3D) | Offline |     [SAM](https://github.com/facebookresearch/segment-anything)     | 18.0 |  32.8 |  56.1 |     --    |     --    |\r\n|   [SAI3D](https://github.com/yd-yin/SAI3D)  | Offline | [SemanticSAM](https://github.com/UX-Decoder/Semantic-SAM) | 30.8 |  50.5 |  70.6 |     --    |     --    |\r\n|   [SAM3D](https://github.com/Pointcept/SegmentAnything3D)  |  Online |     SAM     | 20.6 |  35.7 |  55.5 | 1369+1518 |     --    |\r\n|   ESAM   |  Online |     SAM     | 42.2 |  63.7 |  79.6 |  1369+**80**  |   [model](https://cloud.tsinghua.edu.cn/f/426d6eb693ff4b1fa04b/?dl=1)   |\r\n|  ESAM-E  |  Online |   [FastSAM](https://github.com/CASIA-IVA-Lab/FastSAM)   | **43.4** |  **65.4** |  **80.9** |   **20**+**80**   |   [model](https://cloud.tsinghua.edu.cn/f/7578d7e3d6764f6a93ee/?dl=1)   |\r\n\r\n**Dataset transfer results from ScanNet200 to SceneNN and 3RScan:**\r\n\u003ctable class=\"tg\"\u003e\u003cthead\u003e\r\n  \u003ctr\u003e\r\n    \u003cth class=\"tg-b2st\" rowspan=\"2\"\u003eMethod\u003c/th\u003e\r\n    \u003cth class=\"tg-b2st\" rowspan=\"2\"\u003eType \u003c/th\u003e\r\n    \u003cth class=\"tg-b2st\" colspan=\"3\"\u003eScanNet200--\u0026gt;SceneNN\u003c/th\u003e\r\n    \u003cth class=\"tg-b2st\" colspan=\"3\"\u003eScanNet200--\u0026gt;3RScan\u003c/th\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003cth class=\"tg-wa1i\"\u003eAP\u003c/th\u003e\r\n    \u003cth class=\"tg-wa1i\"\u003eAP@50\u003c/th\u003e\r\n    \u003cth class=\"tg-wa1i\"\u003eAP@25\u003c/th\u003e\r\n    \u003cth class=\"tg-wa1i\"\u003eAP\u003c/th\u003e\r\n    \u003cth class=\"tg-wa1i\"\u003eAP@50\u003c/th\u003e\r\n    \u003cth class=\"tg-wa1i\"\u003eAP@25\u003c/th\u003e\r\n  \u003c/tr\u003e\u003c/thead\u003e\r\n\u003ctbody\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eSAMPro3D\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eOffline\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e12.6\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e25.8\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e53.2\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e3.9\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e8.0\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e21.0\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eSAI3D\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eOffline\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e18.6\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e34.7\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e65.7\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e5.4\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e11.8\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e27.4\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eSAM3D\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eOnline\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e15.1\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e30.0\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e51.8\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e6.2\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e13.0\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e33.9\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eESAM\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eOnline\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e\u003cb\u003e28.8\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e\u003cb\u003e52.2\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e69.3\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e\u003cb\u003e14.1\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e\u003cb\u003e31.2\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e\u003cb\u003e59.6\u003c/b\u003e\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eESAM-E\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eOnline\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e28.6\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e50.4\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e\u003cb\u003e71.0\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e13.9\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e29.4\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e58.8\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n\u003c/tbody\u003e\u003c/table\u003e\r\n\r\n**3D instance segmentation results on ScanNet dataset:**\r\n\u003ctable class=\"tg\"\u003e\u003cthead\u003e\r\n  \u003ctr\u003e\r\n    \u003cth class=\"tg-gabo\" rowspan=\"2\"\u003eMethod\u003c/th\u003e\r\n    \u003cth class=\"tg-gabo\" rowspan=\"2\"\u003eType\u003c/th\u003e\r\n    \u003cth class=\"tg-gabo\" colspan=\"3\"\u003eScanNet\u003c/th\u003e\r\n    \u003cth class=\"tg-gabo\" colspan=\"3\"\u003eSceneNN\u003c/th\u003e\r\n    \u003cth class=\"tg-gabo\" rowspan=\"2\"\u003eFPS\u003c/th\u003e\r\n    \u003cth class=\"tg-gabo\" rowspan=\"2\"\u003eDownload\u003c/th\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003cth class=\"tg-uzvj\"\u003eAP\u003c/th\u003e\r\n    \u003cth class=\"tg-uzvj\"\u003eAP@50\u003c/th\u003e\r\n    \u003cth class=\"tg-uzvj\"\u003eAP@25\u003c/th\u003e\r\n    \u003cth class=\"tg-uzvj\"\u003eAP\u003c/th\u003e\r\n    \u003cth class=\"tg-uzvj\"\u003eAP@50\u003c/th\u003e\r\n    \u003cth class=\"tg-uzvj\"\u003eAP@25\u003c/th\u003e\r\n  \u003c/tr\u003e\u003c/thead\u003e\r\n\u003ctbody\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e\u003ca href=https://github.com/SamsungLabs/td3d\u003eTD3D\u003c/a\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003eoffline\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e46.2\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e71.1\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e81.3\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e\u003ca href=https://github.com/oneformer3d/oneformer3d\u003eOneformer3D\u003c/a\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003eoffline\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e59.3\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e78.8\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e86.7\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e\u003ca href=https://github.com/THU-luvision/INS-Conv\u003eINS-Conv\u003c/a\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003eonline\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e57.4\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e\u003ca href=https://github.com/xuxw98/Online3D\u003eTD3D-MA\u003c/a\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003eonline\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e39.0\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e60.5\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e71.3\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e26.0\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e42.8\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e59.2\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e3.5\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e--\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003eESAM-E\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003eonline\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e41.6\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e60.1\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e75.6\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e27.5\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e48.7\u003c/td\u003e\r\n    \u003ctd class=\"tg-uzvj\"\u003e\u003cb\u003e64.6\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-uzvj\"\u003e\u003cb\u003e10\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-9wq8\"\u003e\u003ca href=https://cloud.tsinghua.edu.cn/f/1eeff1152a5f4d4989da/?dl=1\u003emodel\u003c/a\u003e\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n  \u003ctr\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eESAM-E+FF\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003eonline\u003c/td\u003e\r\n    \u003ctd class=\"tg-wa1i\"\u003e\u003cb\u003e42.6\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-wa1i\"\u003e\u003cb\u003e61.9\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-wa1i\"\u003e\u003cb\u003e77.1\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-wa1i\"\u003e\u003cb\u003e33.3\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-wa1i\"\u003e\u003cb\u003e53.6\u003c/b\u003e\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e62.5\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e9.8\u003c/td\u003e\r\n    \u003ctd class=\"tg-nrix\"\u003e\u003ca href=https://cloud.tsinghua.edu.cn/f/4c2dd1559e854f48be76/?dl=1\u003emodel\u003c/a\u003e\u003c/td\u003e\r\n  \u003c/tr\u003e\r\n\u003c/tbody\u003e\u003c/table\u003e\r\n\r\n**Open-Vocabulary 3D instance segmentation results on ScanNet200 dataset:**\r\n| Method |  AP  | AP@50 | AP@25 |\r\n|:------:|:----:|:-----:|:-----:|\r\n|  SAI3D |  9.6 |  14.7 |  19.0 |\r\n|  ESAM  | **13.7** |  **19.2** |  **23.9** |\r\n\r\n\r\n## TODO List\r\n- [x] Release code and checkpoints.\r\n- [x] Release the demo code to directly run ESAM on streaming RGB-D video.\r\n\r\n## Contributors\r\nBoth students below contributed equally and the order is determined by random draw.\r\n- [Xiuwei Xu](https://xuxw98.github.io/)\r\n- Huangxing Chen\r\n\r\nBoth advised by [Jiwen Lu](https://ivg.au.tsinghua.edu.cn/Jiwen_Lu/).\r\n\r\n## Acknowledgement\r\nWe thank a lot for the flexible codebase of [Oneformer3D](https://github.com/oneformer3d/oneformer3d) and [Online3D](https://github.com/xuxw98/Online3D), as well as the valuable datasets provided by [ScanNet](https://github.com/ScanNet/ScanNet), [SceneNN](https://github.com/hkust-vgd/scenenn) and [3RScan](https://github.com/WaldJohannaU/3RScan).\r\n\r\n\r\n## Citation\r\n```\r\n@article{xu2024esam, \r\n      title={EmbodiedSAM: Online Segment Any 3D Thing in Real Time}, \r\n      author={Xiuwei Xu and Huangxing Chen and Linqing Zhao and Ziwei Wang and Jie Zhou and Jiwen Lu},\r\n      journal={arXiv preprint arXiv:2408.11811},\r\n      year={2024}\r\n}\r\n```\r\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fxuxw98%2FESAM","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fxuxw98%2FESAM","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fxuxw98%2FESAM/lists"}