MotionXperts.github.io/index.html at main · MotionXperts/MotionXperts.github.io · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
<!DOCTYPE html>
<html>
<head>
  <script type="text/x-mathjax-config">
  MathJax.Hub.Config({
    tex2jax: {inlineMath: [['$','$'], ['\\(','\\)']]}
  });
  </script>
  <script src='https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/latest.js?config=TeX-MML-AM_CHTML' async></script>

  <meta charset="utf-8">
  <meta name="description"
        content="CoachMe: Decoding Sport Elements with a Reference-Based Coaching Instruction Generation Model">
  <meta name="keywords" content="Generation Model, Graph Covolutional Network, Language Model, Sports, Coaching Instruction">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>CoachMe</title>

  <!-- Global site tag (gtag.js) - Google Analytics -->
  <script async src="https://www.googletagmanager.com/gtag/js?id=G-PYVRSFMDRL"></script>
  <script>
    window.dataLayer = window.dataLayer || [];

    function gtag() {
      dataLayer.push(arguments);
    }

    gtag('js', new Date());

    gtag('config', 'G-PYVRSFMDRL');
  </script>

  <link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
        rel="stylesheet">

  <link rel="stylesheet" href="./static/css/bulma.min.css">
  <link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
  <link rel="stylesheet" href="./static/css/bulma-slider.min.css">
  <link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
  <link rel="stylesheet"
        href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
  <link rel="stylesheet" href="./static/css/index.css">
  <link rel="icon" href="./static/images/MotionXpert.png">

  <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
  <script defer src="./static/js/fontawesome.all.min.js"></script>
  <script src="./static/js/bulma-carousel.min.js"></script>
  <script src="./static/js/bulma-slider.min.js"></script>
  <script src="./static/js/index.js"></script>
</head>
<body>

<section class="hero">
  <div class="hero-body">
    <div class="container is-max-desktop">
      <div class="columns is-centered">
        <div class="column has-text-centered">
          <h1 class="title is-1 publication-title">CoachMe<br/>
            <p class="title is-3 publication-title">Decoding Sport Elements with a Reference-Based <br/> Coaching
Instruction Generation Model </p>
          </h1>
         <!--  <h1 class="title is-4" style="color: #5c5c5c;">arXiv 2024</h1> -->
          <h1 class="title is-4" style="color: #D21F3C;">ACL 2025</h1>
          <div class="is-size-5 publication-authors">
            <span class="author-block">
              <a href="https://github.com/weihsinyeh">Wei-Hsin Yeh<sup>1,3</sup></a>,</span>
            <span class="author-block">
              <a href="https://github.com/b10815061">Yu-An Su<sup>1</sup></a>,
            </span>
            <span class="author-block">
              <a href="https://github.com/nthuandrew">Chih-Ning Chen<sup>1</sup></a>,
            </span>
            <span class="author-block">
              <a href="https://github.com/allen-sk8">Yi-Hsueh Lin<sup>1,2</sup></a>,
            </span><br>
            <span class="author-block">
              <a href="https://">Calvin Ku<sup>2</sup></a>,
            </span>
            <span class="author-block">
              <a href="https://">Wen-Hsin Chiu<sup>2</sup></a>,
            </span>
            <span class="author-block">
              <a href="https://">Min-Chun Hu<sup>2</sup></a>,
            </span>
            <span class="author-block">
              <a href="https://">Lun-Wei Ku<sup>1</sup></a>,
            </span>
          </div>

          <div class="is-size-5 publication-authors">
            <span class="author-block" style="margin-right: 1em;">Institute of Information Science, Academia Sinica<sup>1</sup></span><br>
            <span class="author-block" style="margin-right: 1em;">National Tsing Hua University<sup>2</sup></span><br>
            <span class="author-block" style="margin-right: 1em;">National Taiwan University<sup>3</sup></span><br>
          </div>
          <div class="is-size-5 publication-authors">
            <span class="author-block" style="margin-right: 1em;">{weihsinyeh168, allen0512911}@gmail.com</span><br>
            <span class="author-block" style="margin-right: 1em;">{yuansu, andrewman71, lwku}@iis.sinica.edu.tw</span><br>
            <span class="author-block" style="margin-right: 1em;"> calvinku@gapp.nthu.edu.tw, whchiu@mx.nthu.edu.tw, anitahu@cs.nthu.edu.tw</span><br>
          </div>
          <div class="column has-text-centered">
            <div class="publication-links">
              <!-- PDF Link. -->
              <span class="link-block">
                <a href="https://aclanthology.org/2025.acl-long.1413/"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="fas fa-file-pdf"></i>
                  </span>
                  <span>Paper</span>
                </a>
              </span>
              <!-- Code Link. -->
              <span class="link-block">
                <a href="https://github.com/MotionXperts/MotionExpert"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="fab fa-github"></i>
                  </span>
                  <span>Code</span>
                  </a>
              </span>
              <span class="link-block">
                <a href="https://github.com/MotionXperts/MotionExpert/tree/main/dataset"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="fas fa-database"></i>
                  </span>
                  <span>Dataset</span>
                  </a>
              </span>
              <span class="link-block">
                <a href="https://arxiv.org/abs/2509.11698"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="ai ai-arxiv"></i>
                  </span>
                  <span>arXiv</span>
                </a>
              </span>
            </div>
          </div>
        </div>
      </div>
    </div>
  </div>
</section>

<section class="hero teaser">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <div class="columns is-centered">
      <div class="column">
        <div class="content">
           <center>
            <h1>
              <strong>CoachMe</strong>
            </h1>
            <div class="content has-text-justified">
            <p>CoachMe aims to democratize access to coaching, helping athletes improve on their own. Users can upload videos of their movements, and CoachMe analyzes the motion to provide precise and actionable instructions.</p>
            <h5>Why is this challenging?</h5>
 If we upload a video to a multimodal model such as GPT, it often produces generic and redundant advice — which is not effective for athletes. The real challenge of the “Motion to Instruction” task is generating feedback that is highly informative and practical.<br><br>
<h5>Our approach</h5>
 CoachMe simulates the way real coaches think. Since collecting large amounts of professional sports data is difficult, we cannot directly train a model to understand what “correct motion” is. Instead, CoachMe analyzes the learner’s motion, compares it with a reference, and generates coaching instructions based on the differences.<br><br>
            </div>
          </center>
          <center>
          <img src="static/images/FS_Demo1.png" class="interpolation-image">
          </center>
          <br>
          <br>
          <br>
          <p>
            <center>
            <img src="static/images/FS_Demo2.png" class="interpolation-image">
          </center>
          <p></p>
          <center>
        <div class="content">
          <img src="static/images/BX_Demo.png"
            class="interpolation-image">
        </div>
        </center>
        </div>
      </div>
      </div>
      <div class="columns is-centered">
       <div class="column is-full-width">
      <br>
      <p>
     <strong>CoachMe</strong> integrates an attention mechanism that focuses on the skeletal graph to generate sports instructions.
     </p><p>
    The attention mechanism is visualized: each video shows a skeletal attention graph highlighting different key joints. We illuminate the three most important joints and the three most crucial relationships between pairs of joints.
     </p><p>
      Each set of images is accompanied by three instructional prompts, each generated by a different model—CoachMe, LLaMA, and GPT-4o—providing predicted instruction based on the motion video.
      </p>
    </div>
    </div>
  </div>
</section>

<section class="section">
  <div class="container is-max-desktop">

    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <h2 class="title is-3">How can an AI model coach intense sport?</h2>
        <h4 class="title is-4">AAAI-25 Educational AI Video Winner</h4>
        <div class="content has-text-justified">
          <div class="video-container" style="position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;">
            <center>
            <iframe width="560" height="315" src="https://www.youtube.com/embed/m7LDiiOyHjQ?si=EVTD0DTr2wmy6Agn" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
            </center>
          </div>
        </div>
      </div>
    </div>

  </div>
</section>

<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered">
      <div class="column is-full-width">
        <h2 class="title is-3">Abstract</h2>
<strong>Motion instruction</strong> is a crucial task that helps athletes refine their technique by analyzing movements and providing corrective guidance. Although recent advances in multimodal models have improved motion understanding, generating precise and sport-specific instruction remains challenging due to the highly domain specific nature of sports and the need for in formative guidance. We propose <strong>CoachMe, a reference-based model that analyzes the differences between a learner’s motion and a reference under temporal and physical aspects.</strong> This approach enables both domain-knowledge learning and the acquisition of a coach-like thinking process that identifies movement errors effectively and provides feedback to explain how to improve. In this paper, we illustrate how CoachMe adapts well to specific sports such as skating and boxing by learning from general movements and then leveraging limited data. Experiments show that CoachMe provides high-quality instructions instead of directions merely in the tone of a coach but without critical information. CoachMe outperforms GPT-4o by 31.6% in G-Eval on figure skating and by 58.3% on boxing. Analysis further confirms that it elaborates on errors and their corresponding improvement methods in the generated instructions.
      </div>
    </div>
  </div>
</section>

<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered">
      <div class="column is-full-width">
        <h2 class="title is-3">Overall Framework</h2>
        <div class="content has-text-justified">
          <p>
            <strong>Overall framework of CoachMe.</strong>
            CoachMe architecture comprises three modules: Concept Difference (Sec. 3.1), Human Pose Perception (Sec. 3.2), and Instruct Motion (Sec. 3.3). Instruct Motion compares the motion Tokenlearner with Tokenref to obtain the difference Tokendiff and take Tokenlearner and Tokendiff as input to the LM to generate instructions.
          </p>
        </div>
      </div>
    </div>

    <div class="column">
        <div class="content">
          <img src="static/images/Framework_CoachMe.png"
            class="interpolation-image">
        </div>
      </div>
</section>
  </div>
</section>

<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered">
      <div class="column is-full-width">
        <h2 class="title is-3">Distribution of Sport Indicators</h2>

      <br>
        <div class="content has-text-justified">
          <p>
              We propose six sport indicators to evaluate Sport Utility,
which is the amount of the information in an instruction: </p>
<p><img src="static/images/error.png"
                class="interpolation-image"> error detection,  <img src="static/images/time.png"
                class="interpolation-image"> time detection,  <img src="static/images/body.png"
                class="interpolation-image"> body part detection,  <img src="static/images/causation.png"
                class="interpolation-image"> causation,  <img src="static/images/method.png"
                class="interpolation-image"> method and  <img src="static/images/coordination.png"
                class="interpolation-image"> coordination. These indicators are designed to assess various aspects of the instructions, ensuring they are comprehensive and effective for coaching purposes.
            </p>
          <p>
            By analyzing the proportion of each sport indicator present in the instructional prompts across the entire dataset (train + test), we obtain the matrix on the far left, where each value represents a proportion.
            </p>
            <center>
            <div class="column">
            <div class="content">
              <img src="static/images/ScoreInGTMatrix.png"
                class="interpolation-image">
            </div>
            </center>
      </div>
            <p>
            We also analyze the distribution of sport indicators predicted in the instructional prompts generated by different models—CoachMe, LLaMa, and GPT-4o—based on videos from the test dataset, and incorporate the G-eval consistency scores, which assess consistency with the ground truth. These analyses result in the three matrices on the right.
            </p>
            <p>
            Each value in these matrices represents the total G-eval score accumulated across all instructional prompts in which the two corresponding sport indicators co-occur, normalized by the total number of prompts multiplied by the maximum possible G-eval score.
          </p>
          <center>
            <div class="column">
            <div class="content">
              <img src="static/images/ScoreInMatrix.png"
                class="interpolation-image">
            </div>
            </center>
        </div>
      </div>
    </div>
    <center>
    <div class="column">
        <div class="content">
          <img src="static/images/Matrix.png"
            class="interpolation-image">
        </div>
      </div>
    </center>
</section>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered">
      <div class="column is-full-width">
        <h2 class="title is-3">Contributions</h2>
        <div class="content has-text-justified">
        <p>1. Proposed a novel task: Motion to Instruction</p>
        <p>2. Released two motion-to-instruction datasets</p>
        <p>3. Introduced sport indicators to evaluate Sport Utility — the amount of useful information in an instruction</p>
        <p>4. Developed CoachMe, which generates accurate instructions, achieves state-of-the-art results, and provides highly informative feedback</p>
        <p>5. Demonstrated that CoachMe adapts effectively to diverse sports with only ~150 training samples, while capturing specific instructional patterns</p>
            </p>
      </div>
        </div>
      </div>
    </div>
</section>
  </div>
</section>
<footer class="footer">
  <div class="container">
    <div class="content has-text-centered">
      <a class="icon-link"
         href="https://aclanthology.org/2025.acl-long.1413/">
        <i class="fas fa-file-pdf"></i>
      </a>
      <a class="icon-link" href="https://github.com/MotionXperts" class="external-link" disabled>
        <i class="fab fa-github"></i>
      </a>
      <a class="icon-link" href="https://arxiv.org/abs/2509.11698" class="external-link" disabled>
        <i class="ai ai-arxiv"></i>
      </a>
    </div>
    <div class="columns is-centered">
      <div class="column is-8">
        <div class="content">
          <p>
            This website is licensed under a <a rel="license"
                                                href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
            Commons Attribution-ShareAlike 4.0 International License</a>.
          </p>
          <p>
            It borrows the source code of <a href="https://github.com/nerfies/nerfies.github.io">this website</a>.
            We would like to thank Utkarsh Sinha and Keunhong Park.
          </p>
        </div>
      </div>
    </div>
  </div>
</footer>

</body>
</html>