-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
495 lines (451 loc) · 21.2 KB
/
index.html
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
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8" />
<!-- Meta tags for social media banners, these should be filled in appropriatly as they are your "business card" -->
<!-- Replace the content tag with appropriate information -->
<meta name="description"
content="End-to-end architectures in autonomous driving (AD) face a significant challenge in interpretability, impeding human-AI trust. Human-friendly natural language has been explored for tasks such as driving explanation and 3D captioning. However, previous works primarily focused on the paradigm of declarative interpretability, where the natural language interpretations are not grounded in the intermediate outputs of AD systems, making the interpretations only declarative. In contrast, aligned interpretability establishes a connection between language and the intermediate outputs of AD systems. Here we introduce Hint-AD, an integrated AD-language system that generates language aligned with the holistic perception-prediction-planning outputs of the AD model. By incorporating the intermediate outputs and a holistic token mixer sub-network for effective feature adaptation, Hint-AD achieves desirable accuracy, achieving state-of-the-art results in driving language tasks including driving explanation, 3D dense captioning, and command prediction. To facilitate further study on driving explanation task on nuScenes, we also introduce a human-labeled dataset, Nu-X. Codes, dataset, and models will be publicly available." />
<meta name="keywords"
content="Hint-AD: Holistically Aligned Interpretability for End-to-End Autonomous Driving" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<title>
Hint-AD | Project Page
</title>
<link rel="icon" type="image/x-icon" href="static/images/favicon.ico" />
<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="stylesheet" href="https://unpkg.com/beerslider/dist/BeerSlider.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="https://documentcloud.adobe.com/view-sdk/main.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">
<h2 class="title is-2 publication-title">
Hint-AD: Holistically Aligned Interpretability for End-to-End Autonomous Driving
</h2>
<p style="color: rgb(169, 60, 60); font-size: 20px;">CoRL 2024</p>
<div class="is-size-5 publication-authors">
<!-- Paper authors -->
<span class="author-block">
<a href="https://robot-k.github.io/" target="_blank">Kairui Ding</a>
<sup>1,3</sup>
,</span>
<span class="author-block">
Boyuan Chen
<sup>1,3</sup>
,</span>
<span class="author-block">
Yuchen Su
<sup>3</sup>
,</span>
<span class="author-block">
<a href="https://c7w.tech/about/" target="_blank">Huan-ang Gao</a>
<sup>1</sup>
,</span>
<span class="author-block">
Bu Jin
<sup>1</sup>
,</span>
<span class="author-block">
Chonghao Sima
<sup>4</sup>
,</span>
<span class="author-block">
Xiaohui Li
<sup>2</sup>
,</span>
<span class="author-block">
Wuqiang Zhang
<sup>2</sup>
,</span>
<span class="author-block">
Paul Barsch
<sup>2</sup>
,</span>
<span class="author-block">
<a href="https://lihongyang.info/" target="_blank">Hongyang Li</a>
<sup>4</sup>
,</span>
<span class="author-block">
<a href="https://sites.google.com/view/fromandto" target="_blank">Hao Zhao</a>
<sup>†1</sup>
</span>
</div>
<!-- a margin of 0.5em -->
<div style="margin: 0.5em;"></div>
<div class="is-size-5 publication-authors">
<span class="author-block is-size-6">
<sup>1</sup> Institute for AI Industry Research (AIR), Tsinghua University <br>
<sup>2</sup> Mercedes-Benz Group China Ltd. <br>
<sup>3</sup> Xingjian College, Tsinghua University <br>
<sup>4</sup> OpenDriveLab, Shanghai AI Lab <br>
<span class="eql-cntrb"><small><sup>†</sup>Indicates Corresponding Author</small></span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- Arxiv PDF link -->
<span class="link-block">
<a href="https://arxiv.org/pdf/2409.06702" target="_blank"
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>
<!-- ArXiv abstract Link -->
<span class="link-block">
<a href="https://arxiv.org/abs/2409.06702" target="_blank"
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>
<!-- Github link -->
<span class="link-block">
<a href="https://github.com/Robot-K/Hint-AD" target="_blank"
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>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- Teaser video-->
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<video poster="" id="tree" autoplay controls muted loop height="100%">
<!-- Your video here -->
<source src="static/video/output.mp4" type="video/mp4" />
</video>
<!-- centering the image -->
<!-- <div class="columns is-centered">
<!-- <div class="column is-four-fifths"> -->
<!-- <div class="publication-video"> -->
<!-- <img src="static/images/Teaser_cs1.jpg" width="100%" /> -->
<!-- </div> -->
<!-- </div> -->
<!-- </div> -->
<!-- <img src="static/images/Teaser_cs1.jpg" width="100%" /> -->
<h2 class="has-text-centered is-size-6">
Demonstration Video of <b>Hint-AD</b>.
</h2>
</div>
</div>
</section>
<!-- End teaser video -->
<!-- Paper abstract -->
<section class="section hero is-light">
<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">Abstract</h2>
<div class="content has-text-justified">
<p>
End-to-end architectures in autonomous driving (AD) face a significant challenge in interpretability, impeding human-AI trust. Human-friendly natural language has been explored for tasks such as driving explanation and 3D captioning. However, previous works primarily focused on the paradigm of declarative interpretability, where the natural language interpretations are not grounded in the intermediate outputs of AD systems, making the interpretations only declarative. In contrast, aligned interpretability establishes a connection between language and the intermediate outputs of AD systems. Here we introduce Hint-AD, an integrated AD-language system that generates language aligned with the holistic perception-prediction-planning outputs of the AD model. By incorporating the intermediate outputs and a holistic token mixer sub-network for effective feature adaptation, Hint-AD achieves desirable accuracy, achieving state-of-the-art results in driving language tasks including driving explanation, 3D dense captioning, and command prediction. To facilitate further study on driving explanation task on nuScenes, we also introduce a human-labeled dataset, Nu-X. Codes, dataset, and models will be publicly available.
</p>
</div>
</div>
</div>
</div>
</section>
<!-- End paper abstract -->
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<h2 class="title is-3">Introduction & Method</h2>
<div class="columns is-centered has-text-centered"
style="width: 100%; display: flex; justify-content: center; align-items: center; flex-direction: column">
<div style="width: 70%; display: flex; justify-content: center; align-items: stretch; flex-direction: row">
<div style="width: 100%;">
<img src="static/images/teaser.png" alt="">
</div>
</div>
<div style="width: 70%;">
<b>Illustration of two paradigms for interpretability</b> of end-to-end autonomous driving (AD) systems through natural language. (a) The <i>declarative</i> interpretability does not utilize intermediate outputs from AD systems, resulting in text that merely justifies the car's driving behavior; (b) <i>Aligned</i> interpretability incorporates intermediate outputs from the AD model to align the generated language with the holistic perception-prediction-planning process.
</div>
</div>
<br /><br />
<div class="columns is-centered has-text-centered"
style="width: 100%; display: flex; justify-content: center; align-items: center; flex-direction: column">
<div style="width: 70%; display: flex; justify-content: center; align-items: stretch; flex-direction: row">
<div style="width: 100%;">
<img src="static/images/method_detailed.png" alt="">
</div>
</div>
<div style="width: 70%;">
<b>Framework of Hint-AD.</b> (a) Hint-AD pipeline illustration. Taking intermediate output tokens from an AD pipeline as input, a language decoder generates natural language responses. A holistic token mixer module is designed to adapt the tokens. (b) Detailed illustration of BEV blocks architecture. (c) A detailed illustration of instance blocks architecture.
</div>
</div>
</div>
</div>
</section>
<section class="hero is-small is-light">
<div class="hero-body">
<div class="container">
<h2 class="title is-3">Dataset</h2>
<div class="columns is-centered has-text-centered"
style="width: 100%; display: flex; justify-content: center; align-items: center; flex-direction: column">
<div style="width: 70%; display: flex; justify-content: center; align-items: stretch; flex-direction: row">
<div style="width: 100%;">
<img src="static/images/dataset.png" alt="">
</div>
</div>
<div style="width: 70%;">
<b>Illustration of Nu-X dataset</b>. Explanation serves as a guide for human learning and understanding. Particularly in the context of end-to-end autonomous driving (AD) systems, human users often seek explanations to bridge the gap between sensor inputs and AD behaviors. Currently, there is no dataset providing such explanations for nuScenes, a widely utilized dataset in AD research. To address this gap and facilitate interpretability-focused research on nuScenes, we introduce Nu-X, a comprehensive, large-scale, human-labeled explanation dataset. Nu-X offers detailed contextual information and diverse linguistic expressions for each of the 34,000 key frames in nuScenes.
</div>
</div>
</div>
</div>
</section>
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<!-- Paper video. -->
<h2 class="title is-3">Results</h2>
<div class="columns is-centered has-text-centered"
style="width: 100%; display: flex; justify-content: center; align-items: center; flex-direction: column">
<div style="width: 65%; display: flex; justify-content: center; align-items: stretch; flex-direction: row">
<div style="width: 100%;">
<img src="static/images/qualitative.png" alt="">
</div>
</div>
<div style="width: 70%;">
<b>Qualitative Results.</b> We present examples of the language output generated by Hint-AD across multiple tasks, including driving explanation, 3D dense captioning, VQA, command prediction, and four categories of alignment tasks. Captions that do not match the ground truth are colored in red.
</div>
<div style="margin: 1em;"></div>
<div style="width: 85%;">
<b>Comparison with baselines.</b> "Inter. outputs" represents intermediate outputs. All methods are adapted for BEV visual representation and employ mixed dataset training. Hint-UniAD and Hint-VAD, as two implementations of Hint-AD on different AD models, outperform baselines across four language tasks in the AD context.
</div>
<!-- Table with training and testing object categories -->
<table border="1">
<tr>
<th rowspan="2"><strong>Input</strong></th>
<th rowspan="2"><strong>Method</strong></th>
<th colspan="5">Nu-X</th>
<th colspan="4">TOD</th>
<th colspan="3">NuScenes-QA</th>
<th rowspan="2">Command Acc.</th>
</tr>
<tr>
<td><strong>C</strong></td>
<td><strong>B</strong></td>
<td><strong>M</strong></td>
<td><strong>R</strong></td>
<td><strong>G</strong></td>
<td><strong>C</strong></td>
<td><strong>B</strong></td>
<td><strong>M</strong></td>
<td><strong>R</strong></td>
<td><strong>H0</strong></td>
<td><strong>H1</strong></td>
<td><strong>All</strong></td>
</tr>
<tr>
<td rowspan="2">Image + 6-shot examples</td>
<td>GPT-4o</td>
<td>19.0</td>
<td>3.95</td>
<td>10.3</td>
<td>24.9</td>
<td>5.22</td>
<td>160.8</td>
<td>50.4</td>
<td>31.6</td>
<td>43.5</td>
<td>42.0</td>
<td>34.7</td>
<td>37.1</td>
<td>75.4</td>
</tr>
<tr>
<td>Gemini 1.5</td>
<td>17.6</td>
<td>3.43</td>
<td>9.3</td>
<td>23.4</td>
<td>5.03</td>
<td>169.7</td>
<td>53.6</td>
<td>33.4</td>
<td>45.9</td>
<td>40.5</td>
<td>32.9</td>
<td>35.4</td>
<td>80.9</td>
</tr>
<tr>
<td rowspan="2">BEV(2D)</td>
<td>ADAPT</td>
<td>17.7</td>
<td>2.06</td>
<td>12.8</td>
<td>27.9</td>
<td>5.79</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>51.0</td>
<td>44.2</td>
<td>46.4</td>
<td>79.3</td>
</tr>
<tr>
<td>BEV+Adapter</td>
<td>18.6</td>
<td>3.47</td>
<td>11.3</td>
<td>24.5</td>
<td>6.27</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>51.8</td>
<td>45.6</td>
<td>47.7</td>
<td>81.1</td>
</tr>
<tr>
<td rowspan="3">BEV(2D) + Bounding Boxes</td>
<td>BEVDet+MCAN</td>
<td>13.2</td>
<td>2.91</td>
<td>10.3</td>
<td>24.5</td>
<td>5.04</td>
<td>104.9</td>
<td>50.1</td>
<td>43.0</td>
<td>68.0</td>
<td><strong>56.2</strong></td>
<td>46.7</td>
<td>49.9</td>
<td>80.7</td>
</tr>
<tr>
<td>Vote2Cap-DETR</td>
<td>15.3</td>
<td>2.61</td>
<td>10.9</td>
<td>24.2</td>
<td>5.33</td>
<td>110.1</td>
<td>48.0</td>
<td>44.4</td>
<td>67.8</td>
<td>51.2</td>
<td>44.9</td>
<td>47.0</td>
<td>76.5</td>
</tr>
<tr>
<td>TOD</td>
<td>14.5</td>
<td>2.45</td>
<td>10.5</td>
<td>23.0</td>
<td>5.10</td>
<td>120.3</td>
<td>51.5</td>
<td>45.1</td>
<td>70.1</td>
<td>53.0</td>
<td>45.1</td>
<td>49.0</td>
<td>78.2</td>
</tr>
<tr>
<td rowspan="2">BEV(2D) + Inter. outputs</td>
<td><strong>Hint-UniAD (Ours)</strong></td>
<td>21.7</td>
<td><strong>4.20</strong></td>
<td>12.7</td>
<td>27.0</td>
<td>7.20</td>
<td><strong>342.6</strong></td>
<td><strong>71.9</strong></td>
<td><strong>48.0</strong></td>
<td><strong>85.4</strong></td>
<td><strong>56.2</strong></td>
<td>47.5</td>
<td>50.4</td>
<td><strong>83.0</strong></td>
</tr>
<tr>
<td><strong>Hint-VAD (Ours)</strong></td>
<td><strong>22.4</strong></td>
<td>4.18</td>
<td><strong>13.2</strong></td>
<td><strong>27.6</strong></td>
<td><strong>7.44</strong></td>
<td>263.7</td>
<td>67.6</td>
<td>47.5</td>
<td>79.4</td>
<td>55.4</td>
<td><strong>48.0</strong></td>
<td><strong>50.5</strong></td>
<td>82.3</td>
</tr>
</table>
<div style="margin: 1em;"></div>
</div>
</div>
</section>
<!-- End youtube video -->
<!--BibTex citation -->
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
If you find our work useful in your research, please consider citing:
<div style="margin: 0.5em;"></div>
<pre><code>@inproceedings{dinghint,
title={Hint-AD: Holistically Aligned Interpretability in End-to-End Autonomous Driving},
author={Ding, Kairui and Chen, Boyuan and Su, Yuchen and Gao, Huan-ang and Jin, Bu and Sima, Chonghao and Li, Xiaohui and Zhang, Wuqiang and Barsch, Paul and Li, Hongyang and others},
booktitle={8th Annual Conference on Robot Learning}
}</code></pre>
</div>
</section>
<!--End BibTex citation -->
<!-- End of Statcounter Code -->
<script src="https://unpkg.com/beerslider/dist/BeerSlider.js"></script>
<script>
new BeerSlider(document.getElementById('slider1'), { start: '40' });
new BeerSlider(document.getElementById('slider2'), { start: '40' });
new BeerSlider(document.getElementById('slider3'), { start: '40' });
new BeerSlider(document.getElementById('slider4'), { start: '40' });
new BeerSlider(document.getElementById('slider5'), { start: '40' });
new BeerSlider(document.getElementById('slider6'), { start: '40' });
new BeerSlider(document.getElementById('slider7'), { start: '40' });
new BeerSlider(document.getElementById('slider8'), { start: '40' });
new BeerSlider(document.getElementById('slider9'), { start: '40' });
new BeerSlider(document.getElementById('slider10'), { start: '40' });
</script>
</body>
</html>