UDAN-CLIP/index.html at main · BRAIN-Lab-AI/UDAN-CLIP · 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
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
<!DOCTYPE html>
<html>

<head>
  <meta charset="utf-8">
  <meta name="description"
    content="UDAN-CLIP: A diffusion-based underwater image enhancement framework with CLIP-guided contrastive learning and spatial attention.">
  <meta name="keywords"
    content="Underwater Image Enhancement, Diffusion Model, CLIP, Vision-Language Learning, Domain Adaptation">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>UDAN-CLIP: Underwater Diffusion Attention Network</title>

  <link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">
  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/bulma/0.9.4/css/bulma.min.css">
  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.5.1/css/all.min.css">
  <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">

  <style>
    /* PaperBanana-inspired styling */
    :root {
      --primary-color: #2c3e50;
      --accent-color: #3498db;
      --light-bg: #f8f9fa;
      --border-radius: 12px;
      --box-shadow: 0 10px 30px rgba(0,0,0,0.1);
    }

    body {
      font-family: 'Google Sans', 'Noto Sans', sans-serif;
      line-height: 1.6;
      color: #2c3e50;
    }

    .hero {
      background: linear-gradient(135deg, #2c3e50 0%, #3498db 100%);
      color: white;
      padding: 4rem 1.5rem;
    }

    .hero .title {
      color: white !important;
      font-size: 2.8rem !important;
      font-weight: 400 !important;
      margin-bottom: 1.5rem !important;
    }

    .hero .subtitle {
      color: rgba(255,255,255,0.9) !important;
      font-size: 1.2rem !important;
    }

    .section {
      padding: 4rem 1.5rem;
    }

    .section-title {
      font-size: 2.2rem;
      font-weight: 400;
      color: var(--primary-color);
      margin-bottom: 2rem;
      text-align: center;
    }

    .content-card {
      background: white;
      border-radius: var(--border-radius);
      padding: 2rem;
      box-shadow: var(--box-shadow);
      margin-bottom: 2rem;
    }

    .publication-authors {
      font-size: 1.2rem;
      margin: 1.5rem 0;
    }

    .author-block {
      margin: 0 0.5rem;
    }

    .publication-links {
      margin: 2rem 0;
    }

    .button.is-rounded {
      border-radius: 30px;
      padding: 0.5rem 1.8rem;
      font-size: 1rem;
      transition: all 0.3s ease;
    }

    .button.is-rounded:hover {
      transform: translateY(-2px);
      box-shadow: 0 5px 15px rgba(0,0,0,0.2);
    }

    .teaser-image {
      border-radius: var(--border-radius);
      box-shadow: var(--box-shadow);
      max-width: 100%;
      height: auto;
      margin: 1rem 0;
    }

    .results-grid {
      display: grid;
      grid-template-columns: repeat(auto-fit, minmax(350px, 1fr));
      gap: 2rem;
      margin: 2rem 0;
    }

    .results-item {
      text-align: center;
      background: white;
      border-radius: var(--border-radius);
      padding: 1.5rem;
      box-shadow: var(--box-shadow);
      transition: transform 0.3s ease;
    }

    .results-item:hover {
      transform: translateY(-5px);
    }

    .results-item img {
      width: 100%;
      border-radius: 8px;
      margin-bottom: 1rem;
    }

    .caption {
      font-size: 0.95rem;
      color: #666;
      margin-top: 0.5rem;
      line-height: 1.4;
    }

    .comparison-row {
      display: flex;
      gap: 2rem;
      margin: 2rem 0;
      flex-wrap: wrap;
      justify-content: center;
    }

    .comparison-col {
      flex: 1;
      min-width: 300px;
      background: white;
      border-radius: var(--border-radius);
      padding: 1.5rem;
      box-shadow: var(--box-shadow);
    }

    .comparison-col img {
      width: 100%;
      border-radius: 8px;
      margin-bottom: 1rem;
    }

    .feature-list {
      list-style: none;
      padding: 0;
    }

    .feature-list li {
      margin: 1rem 0;
      padding-left: 1.5rem;
      position: relative;
    }

    .feature-list li:before {
      content: "▹";
      color: var(--accent-color);
      position: absolute;
      left: 0;
    }

    .abstract-text {
      font-size: 1.1rem;
      text-align: justify;
      max-width: 800px;
      margin: 0 auto;
    }

    .bibtex-code {
      background: var(--light-bg);
      border-radius: var(--border-radius);
      padding: 1.5rem;
      font-family: monospace;
      font-size: 0.9rem;
      overflow-x: auto;
    }

    .footer {
      background: var(--light-bg);
      padding: 3rem 1.5rem;
      margin-top: 2rem;
    }

    .footer a {
      color: var(--primary-color);
      margin: 0 0.5rem;
    }

    .footer a:hover {
      color: var(--accent-color);
    }

    .institution {
      font-size: 1.1rem;
      color: rgba(255,255,255,0.8);
      margin-top: 1rem;
    }

    .full-width-image {
      width: 100%;
      max-width: 1200px;
      margin: 0 auto;
      display: block;
    }

    .table-container {
      background: white;
      border-radius: var(--border-radius);
      padding: 2rem;
      box-shadow: var(--box-shadow);
      margin: 2rem 0;
      overflow-x: auto;
    }

    .table-container img {
      width: 100%;
      height: auto;
      border-radius: 8px;
    }

    @media screen and (max-width: 768px) {
      .hero .title {
        font-size: 2rem !important;
      }
      .section-title {
        font-size: 1.8rem;
      }
      .results-grid {
        grid-template-columns: 1fr;
      }
    }
  </style>
</head>

<body>

  <!-- HERO SECTION -->
  <section class="hero">
    <div class="container">
      <div class="columns is-centered">
        <div class="column is-10 has-text-centered">
          <h1 class="title is-1">
            <b>UDAN-CLIP</b>
          </h1>
          <h2 class="subtitle is-3">
            Underwater Diffusion Attention Network with Contrastive Language-Image Joint Learning
          </h2>

          <div class="publication-authors">
            <span class="author-block">Afrah Shaahid<sup>1</sup>,</span>
            <span class="author-block">Muzammil Behzad<sup>1</sup></span>
          </div>

          <div class="institution">
            <sup>1</sup>King Fahd University of Petroleum and Minerals
          </div>

          <div class="publication-links">
            <span class="link-block">
              <a href="UDANCLIP.pdf" target="_blank" class="button is-rounded is-light">
                <span class="icon"><i class="fas fa-file-pdf"></i></span>
                <span>Paper</span>
              </a>
            </span>
            <span class="link-block">
              <a href="https://arxiv.org/abs/2505.19895" target="_blank" class="button is-rounded is-light">
                <span class="icon"><i class="ai ai-arxiv"></i></span>
                <span>arXiv</span>
              </a>
            </span>
            <span class="link-block">
              <a href="#" target="_blank" class="button is-rounded is-light">
                <span class="icon"><i class="fab fa-github"></i></span>
                <span>Code</span>
              </a>
            </span>
          </div>
        </div>
      </div>
    </div>
  </section>

  <!-- TEASER -->
  <section class="section">
    <div class="container">
      <div class="content has-text-centered">
        <img src="static/images/intro_fig.png" alt="Introduction figure" class="teaser-image" style="max-width: 90%;">
        <p class="caption">
          An overview of UDAN-CLIP addressing underwater image degradation and enhancement challenges.
        </p>
      </div>
    </div>
  </section>

  <!-- ABSTRACT -->
  <section class="section" style="background: var(--light-bg);">
    <div class="container">
      <h2 class="section-title">Abstract</h2>
      <div class="content-card">
        <div class="abstract-text">
          <p>
            Underwater images suffer from color distortion, light absorption,
            and scattering effects that significantly degrade visual quality.
            We propose <strong>UDAN-CLIP</strong>, a diffusion-based underwater image
            enhancement framework augmented with contrastive vision-language
            guidance. Our model integrates domain-adaptive diffusion modeling,
            CLIP-guided semantic alignment, and spatial attention to focus on
            severely degraded regions. Extensive experiments demonstrate
            superior performance over state-of-the-art underwater enhancement
            methods in both quantitative metrics and perceptual quality.
          </p>
        </div>
      </div>
    </div>
  </section>

  <!-- METHOD OVERVIEW -->
  <section class="section">
    <div class="container">
      <h2 class="section-title">Method Overview</h2>
      <div class="content-card">
        <p class="has-text-centered" style="margin-bottom: 2rem;">
          UDAN-CLIP combines diffusion modeling with contrastive language-image supervision to improve semantic consistency during underwater image enhancement.
        </p>

        <div class="comparison-row">
          <div class="comparison-col">
            <img src="static/images/architecture_fig1.png" alt="Architecture overview 1">
            <p class="caption">Overall architecture of UDAN-CLIP</p>
          </div>
          <div class="comparison-col">
            <img src="static/images/architecture_fig2.png" alt="Architecture overview 2">
            <p class="caption">Detailed diffusion module structure</p>
          </div>
        </div>

        <div style="margin-top: 2rem;">
          <ul class="feature-list">
            <li><strong>Domain-Adaptive Diffusion Module:</strong> Learns underwater degradation distributions and progressively restores clean images.</li>
            <li><strong>CLIP-Guided Classifier:</strong> Aligns enhanced images with semantic textual descriptions.</li>
            <li><strong>Spatial Attention Mechanism:</strong> Focuses restoration on heavily degraded regions.</li>
            <li><strong>CLIP-Diffusion Loss:</strong> Encourages semantic consistency during reverse diffusion.</li>
          </ul>
        </div>
      </div>
    </div>
  </section>

  <!-- COMPARISON RESULTS -->
  <section class="section" style="background: var(--light-bg);">
    <div class="container">
      <h2 class="section-title">Comparison with State-of-the-Art</h2>

      <div class="comparison-row">
        <div class="comparison-col">
          <img src="static/images/C60_comparison.png" alt="C60 dataset comparison">
          <p class="caption">Results on C60 dataset</p>
        </div>
        <div class="comparison-col">
          <img src="static/images/T200_comparison.png" alt="T200 dataset comparison">
          <p class="caption">Results on T200 dataset</p>
        </div>
      </div>

      <div class="comparison-row">
        <div class="comparison-col">
          <img src="static/images/Color-Checker_comparison.png" alt="Color-Checker7 dataset comparison">
          <p class="caption">Results on Color-Checker7 dataset</p>
        </div>
        <div class="comparison-col">
          <img src="static/images/heatmap.png" alt="Performance heatmap">
          <p class="caption">Performance heatmap analysis</p>
        </div>
      </div>
    </div>
  </section>

  <!-- QUANTITATIVE RESULTS -->
  <section class="section">
    <div class="container">
      <h2 class="section-title">Quantitative Results</h2>

      <!-- New Quantitative Results Table -->
      <div class="table-container">
        <img src="static/images/results_table.png" alt="Quantitative results table" class="full-width-image">
        <p class="caption has-text-centered">
          Quantitative comparison on T200, Color-Checker7, and C60 datasets. Values show improvement over baseline methods.
        </p>
      </div>

      <div class="results-grid">
        <div class="results-item">
          <img src="static/images/plot1_T200.png" alt="T200 quantitative results">
          <p class="caption">Quantitative comparison on T200 dataset</p>
        </div>
        <div class="results-item">
          <img src="static/images/plot2_Color-Checker7.png" alt="Color-Checker7 quantitative results">
          <p class="caption">Quantitative comparison on Color-Checker7 dataset</p>
        </div>
        <div class="results-item">
          <img src="static/images/plot3_C60.png" alt="C60 quantitative results">
          <p class="caption">Quantitative results on C60 dataset</p>
        </div>
      </div>
    </div>
  </section>

  <!-- DETAILED ANALYSIS -->
  <section class="section" style="background: var(--light-bg);">
    <div class="container">
      <h2 class="section-title">Detailed Analysis</h2>

      <div class="comparison-row">
        <div class="comparison-col">
          <img src="static/images/updated_zoomedin1.png" alt="Zoomed analysis 1">
          <p class="caption">Preservation of fine textures and structural details. Our UDAN-CLIP recovers intricate patterns (e.g., coral textures, fish scales) that are lost or blurred in competing approaches.</p>
        </div>
        <div class="comparison-col">
          <img src="static/images/updated_zoomedin2.png" alt="Zoomed analysis 2">
          <p class="caption">Recovery of fine details in challenging low-light underwater conditions. Our UDAN-CLIP reveals hidden structural elements (e.g., facial features, coin engravings, fish scales, and pool textures) that remain completely obscured in the CLIP-UIE baseline due to severe light absorption and scattering.</p>
        </div>
      </div>
    </div>
  </section>

  <!-- BIBTEX -->
  <section class="section">
    <div class="container">
      <h2 class="section-title">BibTeX</h2>
      <div class="content-card">
        <pre class="bibtex-code"><code>
@article{shaahid2026udanclip,
  title={UDAN-CLIP: Underwater Diffusion Attention Network with Contrastive Language-Image Joint Learning},
  author={Shaahid, Afrah and Behzad, Muzammil},
  journal={arXiv preprint arXiv:2601.xxxxx},
  year={2026}
}
        </code></pre>
      </div>
    </div>
  </section>

  <!-- FOOTER -->
  <footer class="footer">
    <div class="container">
      <div class="content has-text-centered">
        <p>
          <a href="UDANCLIP.pdf" target="_blank"><i class="fas fa-file-pdf fa-2x"></i></a>
          <a href="#" target="_blank" style="margin-left: 20px;"><i class="fab fa-github fa-2x"></i></a>
          <a href="cas-refs.bib" target="_blank" style="margin-left: 20px;"><i class="fas fa-book fa-2x"></i></a>
        </p>
        <p style="margin-top: 2rem; font-size: 0.9rem; color: #666;">
          UDAN-CLIP
        </p>
      </div>
    </div>
  </footer>

</body>
</html>