bag-lab commited on
Commit
026dd84
·
verified ·
1 Parent(s): 91512f0

Upload 3 files

Browse files
Files changed (1) hide show
  1. index.html +804 -18
index.html CHANGED
@@ -1,19 +1,805 @@
1
- <!doctype html>
2
  <html>
3
- <head>
4
- <meta charset="utf-8" />
5
- <meta name="viewport" content="width=device-width" />
6
- <title>My static Space</title>
7
- <link rel="stylesheet" href="style.css" />
8
- </head>
9
- <body>
10
- <div class="card">
11
- <h1>Welcome to your static Space!</h1>
12
- <p>You can modify this app directly by editing <i>index.html</i> in the Files and versions tab.</p>
13
- <p>
14
- Also don't forget to check the
15
- <a href="https://huggingface.co/docs/hub/spaces" target="_blank">Spaces documentation</a>.
16
- </p>
17
- </div>
18
- </body>
19
- </html>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <!DOCTYPE html>
2
  <html>
3
+ <head>
4
+ <meta charset="utf-8">
5
+ <meta name="description"
6
+ content="MARS: A Foundational Map Auto-Regressor .">
7
+ <meta name="keywords" content="Map Generation, Aerial Tracking, Grounding">
8
+ <meta name="viewport" content="width=device-width, initial-scale=1">
9
+ <title>MARS: A Foundational Map Auto-Regressor</title>
10
+
11
+ <!-- Global site tag (gtag.js) - Google Analytics -->
12
+ <script async src="https://www.googletagmanager.com/gtag/js?id=G-PYVRSFMDRL"></script>
13
+ <script>
14
+ window.dataLayer = window.dataLayer || [];
15
+
16
+ function gtag() {
17
+ dataLayer.push(arguments);
18
+ }
19
+
20
+ gtag('js', new Date());
21
+
22
+ gtag('config', 'G-PYVRSFMDRL');
23
+ </script>
24
+
25
+ <link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
26
+ rel="stylesheet">
27
+
28
+ <link rel="stylesheet" href="./static/css/bulma.min.css">
29
+ <link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
30
+ <link rel="stylesheet" href="./static/css/bulma-slider.min.css">
31
+ <link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
32
+ <link rel="stylesheet"
33
+ href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
34
+ <link rel="stylesheet" href="./static/css/index.css">
35
+ <link rel="icon" href="./static/images/favicon.svg">
36
+
37
+ <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
38
+ <script defer src="./static/js/fontawesome.all.min.js"></script>
39
+ <script src="./static/js/bulma-carousel.min.js"></script>
40
+ <script src="./static/js/bulma-slider.min.js"></script>
41
+ <script src="./static/js/index.js"></script>
42
+ </head>
43
+
44
+ <script>
45
+ const slider = document.getElementById("image-slider");
46
+ const img = document.getElementById("slider-image");
47
+
48
+ // List all 4 images in your ./test folder
49
+ const images = [
50
+ "./test/1.png",
51
+ "./test/2.png"
52
+ ];
53
+
54
+ slider.addEventListener("input", () => {
55
+ const index = parseInt(slider.value) - 1;
56
+ img.src = images[index];
57
+ });
58
+ </script>
59
+
60
+
61
+ <body>
62
+
63
+
64
+
65
+ <section class="hero" id="Overview">
66
+ <div class="hero-body">
67
+ <div class="container is-max-desktop">
68
+ <div class="columns is-centered">
69
+ <div class="column has-text-centered">
70
+ <h1 class="title is-1 publication-title">MARS: A Foundational Map Auto-Regressor.</h1>
71
+ <h3 class="is-size-4 has-text-weight-bold" style="color: orange;">
72
+ <!-- Accepted to NeurIPS 2024 Video-Language Modelling Workshop! -->
73
+ </h3>
74
+ <h3 class="is-size-4 has-text-weight-bold" style="color: orange;">
75
+ <!-- Accepted to WACV 2025! -->
76
+ </h3>
77
+ <!-- <div class="is-size-5 publication-authors">
78
+ <span class="author-block">
79
+ <a href="https://sites.google.com/view/rupanjali-kukal/home">Rupanjali Kukal</a>,</span>
80
+ <span class="author-block">
81
+ <a href="https://www.linkedin.com/in/patravali">Jay Patravali</a>,</span>
82
+ <span class="author-block">
83
+ <a href="https://sites.google.com/site/yfuxun">Fuxun Yu</a>,
84
+ </span>
85
+ <span class="author-block">
86
+ <a href="">Simranjit Singh</a>,
87
+ </span>
88
+ <span class="author-block">
89
+ <a href="https://www.linkedin.com/in/nikolaos-karianakis-4016242b?utm_source=share&utm_campaign=share_via&utm_content=profile&utm_medium=ios_app">Nikoloas Karianakis</a>,
90
+ </span>
91
+ <span class="author-block">
92
+ <a href="https://www.linkedin.com/in/rishi-madhok-91663b82?utm_source=share&utm_campaign=share_via&utm_content=profile&utm_medium=ios_app">Rishi Madhok</a>,
93
+ </span>
94
+ </div> -->
95
+
96
+ <div class="is-size-5 publication-authors">
97
+ <span class="author-block"><sup></sup> </span>
98
+ <!-- <span class="author-block"><sup>2</sup>Google Research</span> -->
99
+ </div>
100
+
101
+ <div class="column has-text-centered">
102
+ <div class="publication-links">
103
+ <!-- PDF Link. -->
104
+ <span class="link-block">
105
+ <a href="https://openaccess.thecvf.com/content/WACV2025/papers/Kukal_ClickDescribe_Multimodal_Grounding_and_Tracking_for_Aerial_Objects_WACV_2025_paper.pdf"
106
+ class="external-link button is-normal is-rounded is-dark">
107
+ <span class="icon">
108
+ <i class="fas fa-file-pdf"></i>
109
+ </span>
110
+ <span>Paper</span>
111
+ </a>
112
+ </span>
113
+ <!-- <span class="link-block">
114
+ <a href="https://arxiv.org/abs/2011.12948"
115
+ class="external-link button is-normal is-rounded is-dark">
116
+ <span class="icon">
117
+ <i class="ai ai-arxiv"></i>
118
+ </span>
119
+ <span>arXiv</span>
120
+ </a>
121
+ </span> -->
122
+ <!-- Video Link.
123
+ <span class="link-block">
124
+ <a href="https://www.youtube.com/watch?v=MrKrnHhk8IA"
125
+ class="external-link button is-normal is-rounded is-dark">
126
+ <span class="icon">
127
+ <i class="fab fa-youtube"></i>
128
+ </span>
129
+ <span>Video</span>
130
+ </a>
131
+ </span> -->
132
+ <!-- Code Link. -->
133
+ <span class="link-block">
134
+ <a href=""
135
+ class="external-link button is-normal is-rounded is-dark">
136
+ <span class="icon">
137
+ <i class="fab fa-github"></i>
138
+ </span>
139
+ <span>Code</span>
140
+ </a>
141
+ </span>
142
+ <!-- Dataset Link. -->
143
+ <span class="link-block">
144
+ <a href=""
145
+ class="external-link button is-normal is-rounded is-dark">
146
+ <span class="icon">
147
+ <i class="far fa-images"></i>
148
+ </span>
149
+ <span>Dataset</span>
150
+ </a>
151
+ </div>
152
+ <h5 class="is-size-10 has-text-centered">
153
+ Dataset to be released soon
154
+ </h5>
155
+
156
+ </div>
157
+ </div>
158
+ </div>
159
+ </div>
160
+ </div>
161
+ </section>
162
+
163
+ <nav class="navbar is-fixed-top" role="navigation" aria-label="main navigation" style="background-color: #f5f5f5; box-shadow: 0 2px 4px rgba(0,0,0,0.1);">
164
+ <div class="container">
165
+ <div class="navbar-menu">
166
+ <div class="navbar-start" style="margin: 0 auto;">
167
+ <a class="navbar-item" href="#abstract">Abstract</a>
168
+ <a class="navbar-item" href="#click1-demo">Click 1</a>
169
+ <a class="navbar-item" href="#click2-demo">Click 2</a>
170
+ <a class="navbar-item" href="#click3-demo">Click 3</a>
171
+ <a class="navbar-item" href="#click-jitter">Jitter Analysis</a>
172
+ <a class="navbar-item" href="#interactive-visualizations">Stroke Visualizations</a>
173
+ <a class="navbar-item" href="#failure-cases">Failure Cases</a>
174
+ </div>
175
+ </div>
176
+ </div>
177
+ </nav>
178
+
179
+
180
+
181
+ <section class="hero is-light is-small" id="Inference Results">
182
+ <div class="hero-body">
183
+ <div class="container">
184
+ <div id="results-carousel" class="carousel results-carousel">
185
+ <div class="item item-chair-tp">
186
+ <video poster="" id="chair-tp" autoplay controls muted loop playsinline height="100%">
187
+ <source src="./inference_videos/1.mp4"
188
+ type="video/mp4">
189
+ </video>
190
+ </div>
191
+ <div class="item item-shiba">
192
+ <video poster="" id="shiba" autoplay controls muted loop playsinline height="100%">
193
+ <source src="./inference_videos/2.mp4"
194
+ type="video/mp4">
195
+ </video>
196
+ </div>
197
+ <div class="item item-fullbody">
198
+ <video poster="" id="fullbody" autoplay controls muted loop playsinline height="100%">
199
+ <source src="./inference_videos/3.mp4"
200
+ type="video/mp4">
201
+ </video>
202
+ </div>
203
+ <div class="item item-blueshirt">
204
+ <video poster="" id="blueshirt" autoplay controls muted loop playsinline height="100%">
205
+ <source src="./videos/dtb70/2.mp4"
206
+ type="video/mp4">
207
+ </video>
208
+ </div>
209
+ </div>
210
+ </div>
211
+ </div>
212
+ </section>
213
+
214
+
215
+ <section class="section" id="Abstract">
216
+ <div class="container is-max-desktop">
217
+ <!-- Abstract. -->
218
+ <div class="columns is-centered has-text-centered">
219
+ <div class="column is-four-fifths">
220
+ <h2 class="title is-3">Abstract</h2>
221
+ <div class="content has-text-justified">
222
+ <p>Map generation tasks, featured by extensive non-structural vectorized data (e.g.,
223
+ points, polylines, and polygons), pose significant challenges to common pixelwise generative models. Past works, by segmenting and then performing various
224
+ vectorized post-processing, usually sacrifice accuracy. Motivated by the recent
225
+ huge success of auto-regressive language modeling, we propose the first map
226
+ foundational model: Map Auto-Regressor (MARS), that is capable of generating both multi-polyline road networks and polygon buildings in a unified manner.
227
+ We collected by far the largest multi-class map dataset, MAP-3M, to support the
228
+ robust training. Extensive benchmarks highlight the performance superiority of
229
+ MARS against literature works. Meanwhile, benefited from the auto-regressive
230
+ teaching-forcing based training, we develop the “Chat with MARS” capability
231
+ that enables interactive human-in-the-loop map generation and correction.</p>
232
+
233
+ </div>
234
+ </div>
235
+ </div>
236
+ </div>
237
+ </section>
238
+
239
+
240
+ <section class="section" id="Click 1 Demo">
241
+ <div class="container is-max-desktop">
242
+ <!-- Animation. -->
243
+ <div class="columns is-centered has-text-centered">
244
+ <div class="column is-full-width">
245
+ <h2 class="title is-3">Click 1 Demo</h2>
246
+
247
+ <!-- Interpolating. -->
248
+ <!-- <h3 class="title is-4">Interpolating states</h3> -->
249
+ <div class="content has-text-justified">
250
+ <p>
251
+ This is particularly helpful when a test image is extremely blurry or out-of-domain: once the first
252
+ vertex predicted by MARS is ill-conditioned, due to error accumulation of auto-regressive nature, the whole sequence may suffer from less detections. In such case, SOS chatting can greatly improve
253
+ the full image prediction performance.
254
+ </p>
255
+ </div>
256
+ <div class="columns is-vcentered interpolation-panel">
257
+ <div class="column is-6 has-text-centered">
258
+ <h3 class="subtitle is-5">Original Inference</h3>
259
+ <video poster="" autoplay controls muted loop playsinline style="width: 100%; height: auto;">
260
+ <source src="./click1_videos/29638.mp4" type="video/mp4">
261
+ Your browser does not support the video tag.
262
+ </video>
263
+ </div>
264
+ <div class="column is-6 has-text-centered">
265
+ <h3 class="subtitle is-5">After Click</h3>
266
+ <video poster="" autoplay controls muted loop playsinline style="width: 100%; height: auto;">
267
+ <source src="./click1_videos/29638_click.mp4" type="video/mp4">
268
+ Your browser does not support the video tag.
269
+ </video>
270
+ </div>
271
+ </div>
272
+ <br/>
273
+ </div>
274
+ </div>
275
+ <!--/ Animation. -->
276
+ </div>
277
+ </section>
278
+
279
+
280
+ <section class="section" id="Click 2 Demo">
281
+ <div class="container is-max-desktop">
282
+ <!-- Animation. -->
283
+ <div class="columns is-centered has-text-centered">
284
+ <div class="column is-full-width">
285
+ <h2 class="title is-3">Click 2 Demo</h2>
286
+
287
+ <!-- Interpolating. -->
288
+ <!-- <h3 class="title is-4">Interpolating states</h3> -->
289
+ <div class="content has-text-justified">
290
+ <p>
291
+ Mid-of-sequence (MOS) chatting aims to intercept MARS’s prediction sequence when it drifts from
292
+ the desired trajectory, which is common in vectorized road generation. This is particularly helpful when certain predictions in an image needs to be adjusted.
293
+ </p>
294
+ </div>
295
+ <div class="columns is-vcentered interpolation-panel">
296
+ <div class="column is-6 has-text-centered">
297
+ <h3 class="subtitle is-5">Original Inference</h3>
298
+ <video poster="" autoplay controls muted loop playsinline style="width: 100%; height: auto;">
299
+ <source src="./click2_videos/0.mp4" type="video/mp4">
300
+ Your browser does not support the video tag.
301
+ </video>
302
+ </div>
303
+ <div class="column is-6 has-text-centered">
304
+ <h3 class="subtitle is-5">After Click</h3>
305
+ <video poster="" autoplay controls muted loop playsinline style="width: 100%; height: auto;">
306
+ <source src="./click2_videos/0_click.mp4" type="video/mp4">
307
+ Your browser does not support the video tag.
308
+ </video>
309
+ </div>
310
+ </div>
311
+ <br/>
312
+ </div>
313
+ </div>
314
+ <!--/ Animation. -->
315
+ </div>
316
+ </section>
317
+
318
+
319
+
320
+ <section class="section" id="Click 3 Demo">
321
+ <div class="container is-max-desktop">
322
+ <!-- Animation. -->
323
+ <div class="columns is-centered has-text-centered">
324
+ <div class="column is-full-width">
325
+ <h2 class="title is-3">Click 3 Demo</h2>
326
+
327
+ <!-- Interpolating. -->
328
+ <!-- <h3 class="title is-4">Interpolating states</h3> -->
329
+ <div class="content has-text-justified">
330
+ <p>
331
+ Mid-of-sequence (MOS) chatting aims to intercept MARS’s prediction sequence when it drifts from
332
+ the desired trajectory, which is common in vectorized road generation. This is particularly helpful when certain predictions in an image needs to be adjusted.
333
+ End-of-sequence (EOS) chatting aims to augment MARS’s prediction when there are objects missed
334
+ from the final predictions, which is common for various small map elements.
335
+ </p>
336
+ </div>
337
+ <div class="columns is-vcentered interpolation-panel">
338
+ <div class="column is-6 has-text-centered">
339
+ <h3 class="subtitle is-5">Original Inference</h3>
340
+ <video poster="" autoplay controls muted loop playsinline style="width: 100%; height: auto;">
341
+ <source src="./click3_videos/53819.mp4" type="video/mp4">
342
+ Your browser does not support the video tag.
343
+ </video>
344
+ </div>
345
+ <div class="column is-6 has-text-centered">
346
+ <h3 class="subtitle is-5">After Click</h3>
347
+ <video poster="" autoplay controls muted loop playsinline style="width: 100%; height: auto;">
348
+ <source src="./click3_videos/53819_click.mp4" type="video/mp4">
349
+ Your browser does not support the video tag.
350
+ </video>
351
+ </div>
352
+ </div>
353
+ <br/>
354
+ </div>
355
+ </div>
356
+ <!--/ Animation. -->
357
+ </div>
358
+ </section>
359
+
360
+
361
+
362
+ <!-- CLICK JITTER -->
363
+ <section class="section" id="Click Jitter">
364
+ <div class="container is-max-desktop">
365
+ <div class="columns is-centered has-text-centered">
366
+ <div class="column is-full-width">
367
+ <h2 class="title is-3">Click Jitter Analysis</h2>
368
+
369
+ <div class="content has-text-justified" style="margin-top: 40px;">
370
+ <h3 class="title is-4">Robustness to Click Position Variations</h3>
371
+ <p>The left image shows the ground truth reference. The middle image shows the prediction with correct click position. Use the slider to browse through predictions with different click positions.</p>
372
+ </div>
373
+
374
+ <div class="columns is-vcentered">
375
+ <!-- Left column: Fixed ground truth image -->
376
+ <div class="column is-4 has-text-centered">
377
+ <h4 class="subtitle is-6" style="margin-bottom: 10px;">Ground Truth</h4>
378
+ <img src="./click_jitter/initial.png"
379
+ alt="Ground Truth"
380
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
381
+ </div>
382
+
383
+ <!-- Middle column: Fixed correct click position -->
384
+ <div class="column is-4 has-text-centered">
385
+ <h4 class="subtitle is-6" style="margin-bottom: 10px;">Correct Click Position</h4>
386
+ <img src="./click_jitter/101_77.png"
387
+ alt="Correct Click Position"
388
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
389
+ </div>
390
+
391
+ <!-- Right column: Slider with jitter images -->
392
+ <div class="column is-4 has-text-centered">
393
+ <h4 class="subtitle is-6" style="margin-bottom: 10px;">Click Position Variations</h4>
394
+ <img id="slider-jitter"
395
+ src="./click_jitter/101_64.png"
396
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
397
+ <input class="slider is-fullwidth is-large is-info"
398
+ id="slider-jitter-input"
399
+ type="range" min="1" max="5" value="1" step="1"
400
+ style="margin-top: 20px;">
401
+ <p class="has-text-centered" style="margin-top: 10px;">
402
+ <span id="counter-jitter">Image 1 of 5</span>
403
+ </p>
404
+ </div>
405
+ </div>
406
+ </div>
407
+ </div>
408
+ </div>
409
+ </section>
410
+
411
+ <script>
412
+ document.addEventListener('DOMContentLoaded', function() {
413
+ // Click jitter images (excluding 101_77.png since it's fixed in the middle)
414
+ const jitterImages = [
415
+ "./click_jitter/101_64.png",
416
+ "./click_jitter/101_70.png",
417
+ "./click_jitter/101_85.png",
418
+ "./click_jitter/110_77.png",
419
+ "./click_jitter/90_77.png"
420
+ ];
421
+
422
+ const sliderJitter = document.getElementById('slider-jitter-input');
423
+ const imgJitter = document.getElementById('slider-jitter');
424
+ const counterJitter = document.getElementById('counter-jitter');
425
+
426
+ if (sliderJitter && imgJitter && counterJitter) {
427
+ sliderJitter.max = jitterImages.length;
428
+
429
+ sliderJitter.addEventListener("input", () => {
430
+ const index = parseInt(sliderJitter.value) - 1;
431
+ imgJitter.src = jitterImages[index];
432
+ counterJitter.textContent = `Image ${sliderJitter.value} of ${jitterImages.length}`;
433
+ });
434
+ }
435
+ });
436
+ </script>
437
+
438
+ <!--
439
+ <section class="section" id="BibTeX">
440
+ <div class="container is-max-desktop content">
441
+ <h2 class="title">BibTeX</h2>
442
+ <pre><code>@InProceedings{Kukal_2025_WACV,
443
+ author = {Kukal, Rupanjali and Patravali, Jay and Yu, Fuxun and Singh, Simranjit and Karianakis, Nikolaos and Madhok, Rishi},
444
+ title = {Click\&Describe: Multimodal Grounding and Tracking for Aerial Objects},
445
+ booktitle = {Proceedings of the Winter Conference on Applications of Computer Vision (WACV)},
446
+ month = {February},
447
+ year = {2025},
448
+ pages = {6011-6021}
449
+ }</code></pre>
450
+ </div>
451
+ </section> -->
452
+
453
+
454
+
455
+ <section class="section" id="Stroke Visualizations">
456
+ <div class="container is-max-desktop">
457
+ <div class="columns is-centered has-text-centered">
458
+ <div class="column is-full-width">
459
+ <h2 class="title is-3">Stroke Visualizations</h2>
460
+
461
+ <!-- Intersection Slider -->
462
+ <div class="content has-text-justified" style="margin-top: 40px;">
463
+ <h3 class="title is-4">Intersection Examples</h3>
464
+ <p>Browse through different intersection visualizations.</p>
465
+ </div>
466
+ <div class="columns is-centered">
467
+ <div class="column is-10">
468
+ <img id="slider-intersection"
469
+ src="./stroke_visuals/intersection/19.png"
470
+ style="max-width: 100%; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
471
+ <input class="slider is-fullwidth is-large is-info"
472
+ id="slider-intersection-input"
473
+ type="range" min="1" max="14" value="1" step="1"
474
+ style="margin-top: 20px;">
475
+ <p class="has-text-centered" style="margin-top: 10px;">
476
+ <span id="counter-intersection">Image 1 of 14</span>
477
+ </p>
478
+ </div>
479
+ </div>
480
+
481
+ <!-- Roundabout Slider -->
482
+ <div class="content has-text-justified" style="margin-top: 40px;">
483
+ <h3 class="title is-4">Roundabout Examples</h3>
484
+ <p>Browse through different roundabout visualizations.</p>
485
+ </div>
486
+ <div class="columns is-centered">
487
+ <div class="column is-10">
488
+ <img id="slider-roundabout"
489
+ src="./stroke_visuals/roundabout/65.png"
490
+ style="max-width: 100%; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
491
+ <input class="slider is-fullwidth is-large is-info"
492
+ id="slider-roundabout-input"
493
+ type="range" min="1" max="15" value="1" step="1"
494
+ style="margin-top: 20px;">
495
+ <p class="has-text-centered" style="margin-top: 10px;">
496
+ <span id="counter-roundabout">Image 1 of 15</span>
497
+ </p>
498
+ </div>
499
+ </div>
500
+
501
+ <!-- Splits Slider -->
502
+ <div class="content has-text-justified" style="margin-top: 40px;">
503
+ <h3 class="title is-4">Splits Examples</h3>
504
+ <p>Browse through different road splits visualizations.</p>
505
+ </div>
506
+ <div class="columns is-centered">
507
+ <div class="column is-10">
508
+ <img id="slider-splits"
509
+ src="./stroke_visuals/splits/14.png"
510
+ style="max-width: 100%; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
511
+ <input class="slider is-fullwidth is-large is-info"
512
+ id="slider-splits-input"
513
+ type="range" min="1" max="11" value="1" step="1"
514
+ style="margin-top: 20px;">
515
+ <p class="has-text-centered" style="margin-top: 10px;">
516
+ <span id="counter-splits">Image 1 of 11</span>
517
+ </p>
518
+ </div>
519
+ </div>
520
+
521
+ <!-- T-Junction Slider -->
522
+ <div class="content has-text-justified" style="margin-top: 40px;">
523
+ <h3 class="title is-4">T-Junction Examples</h3>
524
+ <p>Browse through different T-junction visualizations.</p>
525
+ </div>
526
+ <div class="columns is-centered">
527
+ <div class="column is-10">
528
+ <img id="slider-tjunction"
529
+ src="./stroke_visuals/T_Junction/19.png"
530
+ style="max-width: 100%; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
531
+ <input class="slider is-fullwidth is-large is-info"
532
+ id="slider-tjunction-input"
533
+ type="range" min="1" max="10" value="1" step="1"
534
+ style="margin-top: 20px;">
535
+ <p class="has-text-centered" style="margin-top: 10px;">
536
+ <span id="counter-tjunction">Image 1 of 1</span>
537
+ </p>
538
+ </div>
539
+ </div>
540
+
541
+ </div>
542
+ </div>
543
+ </div>
544
+ </section>
545
+
546
+ <script>
547
+ document.addEventListener('DOMContentLoaded', function() {
548
+ // Intersection images
549
+ const intersectionImages = [
550
+ "./stroke_visuals/intersection/19.png",
551
+ "./stroke_visuals/intersection/53.png",
552
+ "./stroke_visuals/intersection/59.png",
553
+ "./stroke_visuals/intersection/119.png",
554
+ "./stroke_visuals/intersection/214.png",
555
+ "./stroke_visuals/intersection/305.png",
556
+ "./stroke_visuals/intersection/382.png",
557
+ "./stroke_visuals/intersection/411.png",
558
+ "./stroke_visuals/intersection/412.png",
559
+ "./stroke_visuals/intersection/676.png",
560
+ "./stroke_visuals/intersection/726.png",
561
+ "./stroke_visuals/intersection/920.png",
562
+ "./stroke_visuals/intersection/1016.png",
563
+ "./stroke_visuals/intersection/1058.png",
564
+ "./stroke_visuals/intersection/1099.png",
565
+ "./stroke_visuals/intersection/1248.png"
566
+ ];
567
+
568
+ // Roundabout images
569
+ const roundaboutImages = [
570
+ "./stroke_visuals/roundabout/65.png",
571
+ "./stroke_visuals/roundabout/83.png",
572
+ "./stroke_visuals/roundabout/108.png",
573
+ "./stroke_visuals/roundabout/115.png",
574
+ "./stroke_visuals/roundabout/173.png",
575
+ "./stroke_visuals/roundabout/187.png",
576
+ "./stroke_visuals/roundabout/217.png",
577
+ "./stroke_visuals/roundabout/224.png",
578
+ "./stroke_visuals/roundabout/356.png",
579
+ "./stroke_visuals/roundabout/400.png",
580
+ "./stroke_visuals/roundabout/413.png",
581
+ "./stroke_visuals/roundabout/420.png",
582
+ "./stroke_visuals/roundabout/528.png",
583
+ "./stroke_visuals/roundabout/612.png",
584
+ "./stroke_visuals/roundabout/688.png",
585
+ "./stroke_visuals/roundabout/798.png"
586
+ ];
587
+
588
+ // Splits images
589
+ const splitsImages = [
590
+ "./stroke_visuals/splits/14.png",
591
+ "./stroke_visuals/splits/54.png",
592
+ "./stroke_visuals/splits/71.png",
593
+ "./stroke_visuals/splits/122.png",
594
+ "./stroke_visuals/splits/135.png",
595
+ "./stroke_visuals/splits/145.png",
596
+ "./stroke_visuals/splits/341.png",
597
+ "./stroke_visuals/splits/386.png",
598
+ "./stroke_visuals/splits/813.png",
599
+ "./stroke_visuals/splits/938.png",
600
+ "./stroke_visuals/splits/953.png"
601
+ ];
602
+
603
+ // T-Junction images (only 1 image visible in folder)
604
+ const tjunctionImages = [
605
+ "./stroke_visuals/T_Junction/21.png",
606
+ "./stroke_visuals/T_Junction/31.png",
607
+ "./stroke_visuals/T_Junction/60.png",
608
+ "./stroke_visuals/T_Junction/254.png",
609
+ "./stroke_visuals/T_Junction/286.png",
610
+ "./stroke_visuals/T_Junction/296.png",
611
+ "./stroke_visuals/T_Junction/302.png",
612
+ "./stroke_visuals/T_Junction/310.png",
613
+ "./stroke_visuals/T_Junction/362.png",
614
+ "./stroke_visuals/T_Junction/749.png"
615
+ ];
616
+
617
+ // Setup slider for Intersection
618
+ setupSlider('intersection', intersectionImages);
619
+ // Setup slider for Roundabout
620
+ setupSlider('roundabout', roundaboutImages);
621
+ // Setup slider for Splits
622
+ setupSlider('splits', splitsImages);
623
+ // Setup slider for T-Junction
624
+ setupSlider('tjunction', tjunctionImages);
625
+
626
+ function setupSlider(name, images) {
627
+ const slider = document.getElementById(`slider-${name}-input`);
628
+ const img = document.getElementById(`slider-${name}`);
629
+ const counter = document.getElementById(`counter-${name}`);
630
+
631
+ slider.max = images.length;
632
+
633
+ slider.addEventListener("input", () => {
634
+ const index = parseInt(slider.value) - 1;
635
+ img.src = images[index];
636
+ counter.textContent = `Image ${slider.value} of ${images.length}`;
637
+ });
638
+ }
639
+ });
640
+ </script>
641
+
642
+
643
+ <section class="section" id="Challenging Cases">
644
+
645
+ <div class="container is-max-desktop">
646
+ <!-- Animation. -->
647
+ <div class="columns is-centered has-text-centered">
648
+ <div class="column is-full-width">
649
+ <h2 class="title is-3"> Challenging Cases</h2>
650
+ <h3 class="title is-4"> Complex Intersection</h3>
651
+
652
+ <!-- Interpolating. -->
653
+ <!-- <h3 class="title is-4">Interpolating states</h3> -->
654
+ <div class="content has-text-justified">
655
+ <p>
656
+ Good Examples.
657
+ </p>
658
+ </div>
659
+ <div class="columns is-vcentered interpolation-panel">
660
+ <div class="column is-4 has-text-centered">
661
+ <!-- <h3 class="subtitle is-5">Original Inference</h3> -->
662
+ <img src="./hardcase_good_visuals/intersection_good/plot_1349291.png"
663
+ alt="Original Inference"
664
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
665
+ </div>
666
+ <div class="column is-4 has-text-centered">
667
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
668
+ <img src="./hardcase_good_visuals/intersection_good/plot_1646496.png"
669
+ alt="After Click"
670
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
671
+ </div>
672
+ <div class="column is-4 has-text-centered">
673
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
674
+ <img src="./hardcase_good_visuals/intersection_good/plot_683728.png"
675
+ alt="After Click"
676
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
677
+ </div>
678
+ </div>
679
+ <div class="content has-text-justified">
680
+ <p>
681
+ Poor Ecamples
682
+ </p>
683
+ </div>
684
+ <div class="columns is-vcentered interpolation-panel">
685
+ <div class="column is-4 has-text-centered">
686
+ <!-- <h3 class="subtitle is-5">Original Inference</h3> -->
687
+ <img src="./failure_case_visuals/complex_intersection/plot_16248.png"
688
+ alt="Original Inference"
689
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
690
+ </div>
691
+ <div class="column is-4 has-text-centered">
692
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
693
+ <img src="./failure_case_visuals/complex_intersection/plot_81558.png"
694
+ alt="After Click"
695
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
696
+ </div>
697
+ <div class="column is-4 has-text-centered">
698
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
699
+ <img src="./failure_case_visuals/complex_intersection/plot_83875.png"
700
+ alt="After Click"
701
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
702
+ </div>
703
+ </div>
704
+ <br/>
705
+ <h3 class="title is-4"> Occlusions</h3>
706
+
707
+ <!-- Interpolating. -->
708
+ <!-- <h3 class="title is-4">Interpolating states</h3> -->
709
+ <div class="content has-text-justified">
710
+ <p>
711
+ Good Examples.
712
+ </p>
713
+ </div>
714
+ <div class="columns is-vcentered interpolation-panel">
715
+ <div class="column is-4 has-text-centered">
716
+ <!-- <h3 class="subtitle is-5">Original Inference</h3> -->
717
+ <img src="./hardcase_good_visuals/occluded_good/plot_292951.png"
718
+ alt="Original Inference"
719
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
720
+ </div>
721
+ <div class="column is-4 has-text-centered">
722
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
723
+ <img src="./hardcase_good_visuals/occluded_good/plot_424350.png"
724
+ alt="After Click"
725
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
726
+ </div>
727
+ <div class="column is-4 has-text-centered">
728
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
729
+ <img src="./hardcase_good_visuals/occluded_good/plot_483392.png"
730
+ alt="After Click"
731
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
732
+ </div>
733
+ </div>
734
+
735
+ <div class="content has-text-justified">
736
+ <p>
737
+ Bad Examples.
738
+ </p>
739
+ </div>
740
+ <div class="columns is-vcentered interpolation-panel">
741
+ <div class="column is-4 has-text-centered">
742
+ <!-- <h3 class="subtitle is-5">Original Inference</h3> -->
743
+ <img src="./failure_case_visuals/occluded/plot_113602.png"
744
+ alt="Original Inference"
745
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
746
+ </div>
747
+ <div class="column is-4 has-text-centered">
748
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
749
+ <img src="./failure_case_visuals/occluded/plot_124720.png"
750
+ alt="After Click"
751
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
752
+ </div>
753
+ <div class="column is-4 has-text-centered">
754
+ <!-- <h3 class="subtitle is-5">After Click</h3> -->
755
+ <img src="./failure_case_visuals/occluded/plot_13641.png"
756
+ alt="After Click"
757
+ style="width: 100%; height: auto; border-radius: 8px; box-shadow: 0 4px 8px rgba(0,0,0,0.1);">
758
+ </div>
759
+ </div>
760
+ <br/>
761
+ <!-- Interpolating. -->
762
+ <!-- <h3 class="title is-4">Interpolating states</h3> -->
763
+ <br/>
764
+ </div>
765
+ </div>
766
+ <!--/ Animation. -->
767
+ </div>
768
+ </section>
769
+
770
+
771
+ <footer class="footer">
772
+ <div class="container">
773
+ <div class="content has-text-centered">
774
+ <a class="icon-link"
775
+ href="./static/videos/nerfies_paper.pdf">
776
+ <i class="fas fa-file-pdf"></i>
777
+ </a>
778
+ <a class="icon-link" href="https://github.com/keunhong" class="external-link" disabled>
779
+ <i class="fab fa-github"></i>
780
+ </a>
781
+ </div>
782
+ <div class="columns is-centered">
783
+ <div class="column is-8">
784
+ <div class="content">
785
+ <p>
786
+ This website is licensed under a <a rel="license"
787
+ href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
788
+ Commons Attribution-ShareAlike 4.0 International License</a>.
789
+ </p>
790
+ <p>
791
+ If you use the source code of this website <a
792
+ href="https://github.com/RupanjaliKukal/Click-Describe-Multimodal-Grounding-and-Tracking-for-Aerial-Objects.git">source code</a>, please also link back <a
793
+ href="https://github.com/nerfies/nerfies.github.io"> Nerfies source code</a> in your footer.
794
+ </p>
795
+ </div>
796
+ </div>
797
+ </div>
798
+ </div>
799
+ </footer>
800
+
801
+
802
+
803
+
804
+ </body>
805
+ </html>