Hugging Face Daily Papers · August 17, 2026 · 3 min read

HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark

Mirrored from Hugging Face Daily Papers for archival readability. Support the source by reading on the original site.

Like Read original ↗

We introduce HumanTracker, a comprehensive and human-aligned benchmark for humanoid motion tracking, together with HumanScore, a preference-aligned metric that better reflects human judgments. Our goal is to move beyond simple kinematic errors and evaluate what truly matters—stable, natural, and physically plausible motion.</p>\n","updatedAt":"2026-08-17T03:09:07.024Z","author":{"_id":"63c3e8abc7d7f4c63a515a02","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63c3e8abc7d7f4c63a515a02/npMHnVP2hHLbvoUGe7C4O.jpeg","fullname":"Zekun Qi","name":"qizekun","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":8,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8153523802757263},"editors":["qizekun"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/63c3e8abc7d7f4c63a515a02/npMHnVP2hHLbvoUGe7C4O.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.13555","authors":[{"_id":"6a7fe4aeb601d59c65281058","name":"Dairu Liu","hidden":false},{"_id":"6a7fe4aeb601d59c65281059","name":"Zekun Qi","hidden":false},{"_id":"6a7fe4aeb601d59c6528105a","name":"Jiayu Zeng","hidden":false},{"_id":"6a7fe4aeb601d59c6528105b","name":"Ruixi Yu","hidden":false},{"_id":"6a7fe4aeb601d59c6528105c","name":"Yu Guan","hidden":false},{"_id":"6a7fe4aeb601d59c6528105d","name":"Yintianrun Zhang","hidden":false},{"_id":"6a7fe4aeb601d59c6528105e","name":"Xuchuan Chen","hidden":false},{"_id":"6a7fe4aeb601d59c6528105f","name":"Sikai Liang","hidden":false},{"_id":"6a7fe4aeb601d59c65281060","name":"Zekai Li","hidden":false},{"_id":"6a7fe4aeb601d59c65281061","name":"Chenghuai Lin","hidden":false},{"_id":"6a7fe4aeb601d59c65281062","name":"Xinqiang Yu","hidden":false},{"_id":"6a7fe4aeb601d59c65281063","name":"Wenyao Zhang","hidden":false},{"_id":"6a7fe4aeb601d59c65281064","name":"He Wang","hidden":false},{"_id":"6a7fe4aeb601d59c65281065","name":"Li Yi","hidden":false}],"publishedAt":"2026-08-13T00:00:00.000Z","submittedOnDailyAt":"2026-08-17T00:00:00.000Z","title":"HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark","submittedOnDailyBy":{"_id":"63c3e8abc7d7f4c63a515a02","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63c3e8abc7d7f4c63a515a02/npMHnVP2hHLbvoUGe7C4O.jpeg","isPro":false,"fullname":"Zekun Qi","user":"qizekun","type":"user","name":"qizekun"},"summary":"Humanoid motion tracking is central to teleoperation and whole-body imitation, yet evaluation often disagrees with what people perceive in videos. Kinematic errors average per-frame pose differences but miss the physical artifacts that matter most, particularly unstable support and incorrect contacts such as foot skating and mistimed touch-downs. Meanwhile, widely used test suites are small and lack the diversity needed to stress contact-rich, long-horizon behaviors. We introduce HumanTracker to make humanoid tracking evaluation both perceptually aligned and scalable. The HumanTracker benchmark contains approximately 153 hours of optical motion trajectories from multiple professional performers, organized into four motion families with text labels for fine-grained diagnosis. We further propose HumanScore, a preference-aligned metric trained on 12K motion pairs containing 24K motions. Across representative state-of-the-art trackers, HumanScore better predicts human preferences and reveals contact and stability failures that kinematic metrics often miss.","upvotes":13,"discussionId":"6a7fe4afb601d59c65281066","projectPage":"https://dairuliu.github.io/humantracker/","githubRepo":"https://github.com/GalaxyGeneralRobotics/HumanTracker","githubRepoAddedBy":"user","ai_summary":"HumanTracker introduces a large-scale benchmark and preference-aligned metric to evaluate humanoid motion tracking based on perceptual quality and physical contact stability.","ai_keywords":["humanoid motion tracking","kinematic errors","HumanTracker","optical motion trajectories","HumanScore","preference-aligned metric","contact-rich behaviors"],"ai_summary_model":"thinkingmachines/Inkling-Small","githubStars":20,"organization":{"_id":"6842da337b70d258bfb51b79","name":"GalaxyGeneralRobotics","fullname":"galbot","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/6842d3b4297a588f54e85f84/q9PZ3SXbzzluHhhMjLD-3.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"63c3e8abc7d7f4c63a515a02","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63c3e8abc7d7f4c63a515a02/npMHnVP2hHLbvoUGe7C4O.jpeg","isPro":false,"fullname":"Zekun Qi","user":"qizekun","type":"user"},{"_id":"683ff8b1f9f5cf052b4db216","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/ugYm9NE_FIGM2Wgp98UQ7.png","isPro":false,"fullname":"jiaqima","user":"jiaqi-ma","type":"user"},{"_id":"683ff9810278714998705654","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/xhvcLacOLbx27WiHY0BZf.png","isPro":false,"fullname":"chenxin ding","user":"chenxin-ding","type":"user"},{"_id":"683ff8138c64413c0d7bfbba","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/ZmBD-cyEuag_Y6JHrd5Il.png","isPro":false,"fullname":"Xukun Cai","user":"ikun-kun","type":"user"},{"_id":"683ffeb9967d02230641c8ad","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/z2WZJkllrCKTCliLEMrr-.png","isPro":false,"fullname":"linkai wang","user":"linkai-wang","type":"user"},{"_id":"68cb1b1ff283dbb922dc8e9c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/zylXL3XGKy5E8tISylviX.png","isPro":false,"fullname":"Yangbowen","user":"Bowen-Yang","type":"user"},{"_id":"689cb792f522165a63e55e4f","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/689cb792f522165a63e55e4f/LIQv_bkx7rqZLax8CAuyV.jpeg","isPro":false,"fullname":"Haiquan Lu","user":"haiquanlu","type":"user"},{"_id":"66839c7f7b0920f40d7ff3e0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/66839c7f7b0920f40d7ff3e0/nuOh5acjmb9QZK3wNQxt1.jpeg","isPro":false,"fullname":"Dairu Liu","user":"dairuliu","type":"user"},{"_id":"68b5b682fbd18ef2d9c011fb","avatarUrl":"/avatars/1f8c7738001da98db5add914dca82fe4.svg","isPro":false,"fullname":"Chenghuai Lin","user":"chenghuailin","type":"user"},{"_id":"65f9533b136fb8ddbd14e1fa","avatarUrl":"/avatars/d88f75da0448093ccd1babba2a37d73f.svg","isPro":false,"fullname":"Zhang","user":"WenyaoZhang","type":"user"},{"_id":"6a827f32d45e5232172d6122","avatarUrl":"/avatars/15e3e0f7ce89656121de561397da01d8.svg","isPro":false,"fullname":"Yaxuan Song","user":"Yaxuan-Song","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"6842da337b70d258bfb51b79","name":"GalaxyGeneralRobotics","fullname":"galbot","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/6842d3b4297a588f54e85f84/q9PZ3SXbzzluHhhMjLD-3.jpeg"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.13555.md","query":{}}">

Papers

arxiv:2608.13555

HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark

Published on Aug 13

· Submitted by

Zekun Qi on Aug 17

galbot

Upvote

Authors:

Abstract

HumanTracker introduces a large-scale benchmark and preference-aligned metric to evaluate humanoid motion tracking based on perceptual quality and physical contact stability.

Generated by thinkingmachines/Inkling-Small

Humanoid motion tracking is central to teleoperation and whole-body imitation, yet evaluation often disagrees with what people perceive in videos. Kinematic errors average per-frame pose differences but miss the physical artifacts that matter most, particularly unstable support and incorrect contacts such as foot skating and mistimed touch-downs. Meanwhile, widely used test suites are small and lack the diversity needed to stress contact-rich, long-horizon behaviors. We introduce HumanTracker to make humanoid tracking evaluation both perceptually aligned and scalable. The HumanTracker benchmark contains approximately 153 hours of optical motion trajectories from multiple professional performers, organized into four motion families with text labels for fine-grained diagnosis. We further propose HumanScore, a preference-aligned metric trained on 12K motion pairs containing 24K motions. Across representative state-of-the-art trackers, HumanScore better predicts human preferences and reveals contact and stability failures that kinematic metrics often miss.