Omnieraser/index.html at main · PRIS-CV/Omnieraser · 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
478
479
480
481
482
483
484
485
486
487
488
489
490
491
<!DOCTYPE html>
<html lang="en">

<link rel="icon" href="static\images\logo_transparent.png" type="image/x-icon" />

<head>
  <meta charset="utf-8">
  <meta name="description"
    content="Discover the OmniEraser, a novel method that seamlessly removes objects and their visual effects">
  <meta property="og:title"
    content="OmniEraser: Remove Objects and Their Effects in Images with Paired Video-Frame Data">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>OmniEraser</title>
  <!-- Fonts and Libraries -->
  <link href="https://fonts.googleapis.com/css2?family=Roboto:wght@400;500;700&display=swap" rel="stylesheet">
  <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bootstrap@5.3.0/dist/css/bootstrap.min.css">
  <link rel="stylesheet" href="https://unpkg.com/aos@2.3.1/dist/aos.css">

  <!-- Custom Styles -->
  <style>
    body {
      font-family: 'Roboto', sans-serif;
      background: linear-gradient(to bottom, #f7f8fc, #e9ecf7);
      color: #333;
    }

    .hero-header {
      background: linear-gradient(135deg, #00c6ff, #0072ff);
      color: #fff;
      padding: 3rem 1rem;
      text-align: center;
      border-radius: 1rem;
      box-shadow: 0 4px 15px rgba(0, 0, 0, 0.2);
    }

    h1,
    h2 {
      font-weight: 700;
    }

    .image-grid {
      display: grid;
      grid-template-columns: repeat(auto-fit, minmax(250px, 1fr));
      gap: 1rem;
      margin-top: 2rem;
    }

    .image-container {
      position: relative;
      overflow: hidden;
      cursor: pointer;
      border-radius: 15px;
      transition: transform 0.3s ease, box-shadow 0.3s ease;
    }

    .image-container:hover {
      transform: scale(1.05);
      box-shadow: 0 10px 25px rgba(0, 0, 0, 0.2);
    }

    .image-container img {
      width: 100%;
      height: 100%;
      object-fit: cover;
      transition: opacity 0.5s ease;
    }

    .image-container .alternate {
      position: absolute;
      top: 0;
      left: 0;
      opacity: 0;
      width: 100%;
      height: 100%;
      object-fit: cover;
    }

    .image-container:active .alternate {
      opacity: 1;
    }

    .image-container:active .original {
      opacity: 0;
    }

    .citation-box {
      background-color: #f8f9fa;
      border-left: 4px solid #007bff;
      overflow-x: auto;
    }

    .copy-btn {
      transition: all 0.3s;
    }

    .copy-btn:hover {
      transform: translateY(-2px);
      box-shadow: 0 5px 15px rgba(0, 0, 0, 0.1);
    }

    .card {
      transition: transform 0.3s, box-shadow 0.3s;
    }

    .card:hover {
      transform: translateY(-5px);
      box-shadow: 0 10px 20px rgba(0, 0, 0, 0.15) !important;
    }
  </style>
</head>

<body>

  <!-- Header Section -->
  <section class="hero-header">
    <div class="logo" style="text-align: center;" data-aos="fade-up" data-aos-delay="100">
      <a href="index.html">
        <img src="static\images\logo_white.png", alt="OmniEraser Logo" width="100" height="100"
          style="margin-bottom: 5px; border-radius: 10px;">
      </a>
    </div>

    <div class="container">
      <h1 data-aos="fade-up" data-aos-delay="200">
        OmniEraser: Remove Objects and Their Effects in Images with Paired Video-Frame Data
      </h1>
      <p data-aos="fade-up" data-aos-delay="300">
        Seamlessly remove objects and their visual effects with innovative paired video-frame data-based solutions.
      </p>
      <!-- Buttons for Code, Dataset, Demo -->
      <div class="d-flex justify-content-center mt-4" data-aos="fade-up" data-aos-delay="400">

        <!-- Arxiv Button -->
        <a href="https://arxiv.org/abs/2501.07397" target="_blank" class="btn btn-dark btn-lg mx-2">
          <span class="icon">
            <img src="static/images/arxiv-logo-one-color-white.svg" width="25" alt="ArXiv">
          </span>
          Paper
        </a>
        <!-- Code Button -->
        <a href="https://github.com/PRIS-CV/Omnieraser" target="_blank" class="btn btn-dark btn-lg mx-2">
          <span class="icon">
            <img src="static/images/gitt.png" width="25" alt="GitHub">
          </span>
          Code
        </a>
        <!-- Dataset Button (coming soon) -->
        <a href="#" target="_blank" class="btn btn-dark btn-lg mx-2" disabled>
          <span class="icon">
            <img src="static/images/hf-logo-pirate.png" width="25" alt="Dataset">
          </span>
          Dataset (Coming soon)
        </a>
        <!-- Demo Button -->
        <a href="https://huggingface.co/spaces/theSure/Omnieraser" target="_blank" class="btn btn-dark btn-lg mx-2" disabled>
          <span class="icon">
            <img src="static/images/output.png" width="25" alt="Demo">
          </span>
          Demo
        </a>

      </div>
    </div>
  </section>

  <!-- Abstract and Authors Section -->
  <section class="py-5 text-center bg-light" data-aos="fade-up" data-aos-delay="400">
    <div class="container is-max-desktop">
      <div class="columns is-centered">
        <div class="column has-text-centered">
          <h2 class="subtitle is-size-5">Authors</h2>
          <div class="is-size-5 publication-authors">
            <span class="author-block">
              <a href="https://scholar.google.com/citations?hl=zh-CN&user=PbO7acIAAAAJ" target="_blank">Runpu
                Wei</a><sup>* 1</sup>,</span>
            <span class="author-block">
              <a href="https://scholar.google.com/citations?hl=zh-CN&user=fFoWQfsAAAAJ" target="_blank">Zijin
                Yin</a><sup>* 1</sup>,</span>
            <span class="author-block">
              <a href="https://scholar.google.com/citations?hl=zh-CN&user=K6rj4gMAAAAJ" target="_blank">Shuo
                Zhang</a><sup>* 1</sup>,</span>
            <span class="author-block">
              <a href="https://dblp.org/search?q=LanXiang+Zhou" target="_blank">LanXiang Zhou</a><sup>2</sup>,</span>
            <span class="author-block">
              <a href="https://github.com/Wesley8170" target="_blank">Xueyi Wang</a><sup>1</sup>,</span>
            <span class="author-block">
              <a href="https://dblp.org/pid/295/8850.html" target="_blank">Chao Ban</a><sup>2</sup>,</span>
            <span class="author-block">
              <a href="https://github.com/caotianwei" target="_blank">Tianwei Cao</a><sup>1</sup>,</span>
            <span class="author-block">
              <a href="https://scholar.google.com/citations?hl=en&user=kYnJS5EAAAAJ" target="_blank">Hao
                Sun</a><sup>2</sup>,</span>
            <span class="author-block">
              <a href="https://dblp.org/pid/348/6925.html" target="_blank">Zhongjiang He</a><sup>2</sup>,</span>
            <span class="author-block">
              <a href="https://scholar.google.com/citations?hl=en&user=dmlkJR4AAAAJ" target="_blank">Kongming
                Liang</a><sup>1</sup>,</span>
            <span class="author-block">
              <a href="https://scholar.google.com/citations?user=5GAAs7IAAAAJ&hl=en" target="_blank">Zhanyu
                Ma</a><sup>1</sup>,</span>
          </div>
        </div>
        <div class="is-size-5 publication-authors">
          <span class="author-block"><sup>1</sup> PRIS, Beijing University of Posts and Telecommunications</span><br>
          <span class="author-block"></span><sup>2</sup> China Telecom Artificial Intelligence Technology
          Co.Ltd</span><br>
          <span class="eql-cntrb"><small><sup>*</sup>Indicates Equal Contribution</small></span>
        </div>
      </div>
    </div>
    </div>
  </section>

  <section class="py-5 text-center" data-aos="fade-up" data-aos-delay="400">
    <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>
              Inpainting algorithms have achieved remarkable progress in removing objects from images, yet still face
              two challenges: 1) struggle to handle the object's visual effects such as shadow and reflection; 2) easily
              generate shape-like artifacts and unintended content. In this paper, we propose Video4Removal, a
              large-scale dataset comprising over 100,000 high-quality samples with realistic object shadows and
              reflections. By constructing object-background pairs from video frames with off-the-shelf vision models,
              the labor costs of data acquisition can be significantly reduced. To avoid generating shape-like artifacts
              and unintended content, we propose Object-Background Guidance, an elaborated paradigm that takes both the
              foreground object and background images. It can guide the diffusion process to harness richer contextual
              information. Based on the above two designs, we present OmniEraser, a novel method that seamlessly removes
              objects and their visual effects using only object masks as input. Extensive experiments show that
              OmniEraser significantly outperforms previous methods, particularly in complex in-the-wild scenes. And it
              also exhibits a strong generalization ability in anime-style images.
            </p>
          </div>
        </div>
      </div>
    </div>
  </section>


  <!-- Visualization Results Section -->
  <section class="py-5 bg-light">
    <div class="container">
      <h2 class="text-center mb-4" data-aos="fade-up" data-aos-delay="400">Visualization Results</h2>
      <p class="text-center" style="color: #0072ff; font-weight: bold;" data-aos="fade-up" data-aos-delay="400">Press
        any image to see the transformation.</p>
      <div class="image-grid" data-aos="fade-up" data-aos-delay="400">
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/0f900fe8-6eab-4f85-8121-29cac9509b94.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/0f900fe8-6eab-4f85-8121-29cac9509b94.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/198b2f55-ccaa-4df4-8216-aab273c8570e.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/198b2f55-ccaa-4df4-8216-aab273c8570e.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/3c43156c-2b44-4ebf-9c47-7707ec60b166.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/3c43156c-2b44-4ebf-9c47-7707ec60b166.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/73ee07ec-7c41-4aef-9769-62213fa00a69.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/73ee07ec-7c41-4aef-9769-62213fa00a69.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/969a07ef-1d5c-42b5-8938-914e3a021781.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/969a07ef-1d5c-42b5-8938-914e3a021781.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/5792ecdb-87b2-4289-ac62-3711063b0cef.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/5792ecdb-87b2-4289-ac62-3711063b0cef.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/ba066105-7fcc-44b5-a20e-7443602d7bb9.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/ba066105-7fcc-44b5-a20e-7443602d7bb9.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/f948ed41-d1f9-4a3b-b40e-9d8c1a9b7c08.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/f948ed41-d1f9-4a3b-b40e-9d8c1a9b7c08.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/5772d5b3-dc43-4b96-ad51-57009287c21c.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/5772d5b3-dc43-4b96-ad51-57009287c21c.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/91105b9a-ca6c-4a8c-a2f7-ea7e121a25e1.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/91105b9a-ca6c-4a8c-a2f7-ea7e121a25e1.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/f8f995cd-335e-481b-a2fe-63389e73e369.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/f8f995cd-335e-481b-a2fe-63389e73e369.png"
            alt="Processed Image">
        </div>
        <div class="image-container">
          <img loading="lazy" class="original" src="static/images/input/67ae0e99-1a2d-488d-b738-24ecf3f45c9a.png" alt="Original Image">
          <img loading="lazy" class="alternate" src="static/images/output/67ae0e99-1a2d-488d-b738-24ecf3f45c9a.png"
            alt="Processed Image">
        </div>
      </div>
    </div>
  </section>

  <!-- Video4Removal Section -->
  <section class="py-5" data-aos="fade-up" data-aos-delay="400">
    <div class="container text-center">
      <h2 class="mb-4" data-aos="fade-up">Video4Removal</h2>
      <p class="mb-4" data-aos="fade-up" data-aos-delay="400">
        We introduce <strong>Video4Removal</strong>, a new photorealistic dataset for object removal, containing over
        10,000 image pairs derived from real-world video recordings.
      </p>

      <!-- Carousel for Video4Removal -->
      <div id="video4removalCarousel" class="carousel slide" data-bs-ride="carousel" data-aos="fade-up" data-aos-delay="400">
        <div class="carousel-inner">
          <div class="carousel-item active">
            <img loading="lazy" src="static/images/sup_vis_video4removal_00(1).png" class="d-block w-100" alt="Image 1">
          </div>
          <div class="carousel-item">
            <img loading="lazy" src="static/images/sup_vis_video4removal_00(2).png" class="d-block w-100" alt="Image 2">
          </div>
          <div class="carousel-item">
            <img loading="lazy" src="static/images/sup_vis_video4removal_00(3).png" class="d-block w-100" alt="Image 3">
          </div>
          <div class="carousel-item">
            <img loading="lazy" src="static/images/sup_vis_video4removal_00(4).png" class="d-block w-100" alt="Image 4">
          </div>
        </div>
        <button class="carousel-control-prev" type="button" data-bs-target="#video4removalCarousel"
          data-bs-slide="prev">
          <span class="carousel-control-prev-icon" aria-hidden="true"></span>
          <span class="visually-hidden">Previous</span>
        </button>
        <button class="carousel-control-next" type="button" data-bs-target="#video4removalCarousel"
          data-bs-slide="next">
          <span class="carousel-control-next-icon" aria-hidden="true"></span>
          <span class="visually-hidden">Next</span>
        </button>
      </div>

    </div>
  </section>

  <!-- RemovalBench Section -->
  <section class="py-5 bg-light" data-aos="fade-up" data-aos-delay="400">
    <div class="container text-center">
      <h2 class="mb-4" data-aos="fade-up">RemovalBench</h2>
      <p class="mb-4" data-aos="fade-up" data-aos-delay="400">
        We introduce a new benchmark <strong>RemovalBench</strong>, which contains 70 pairs of carefully crafted object
        and ground-truth examples.
      </p>

      <!-- Carousel for RemovalBench -->
      <div id="removalBenchCarousel" class="carousel slide" data-bs-ride="carousel" data-aos="fade-up" data-aos-delay="400">
        <div class="carousel-inner">
          <div class="carousel-item active">
            <img src="static/images/sup_vis_removal_benchmark_00(1).png" class="d-block w-100" alt="Image 1">
          </div>
          <div class="carousel-item">
            <img src="static/images/sup_vis_removal_benchmark_00(2).png" class="d-block w-100" alt="Image 2">
          </div>
          <div class="carousel-item">
            <img src="static/images/sup_vis_removal_benchmark_00(3).png" class="d-block w-100" alt="Image 3">
          </div>
          <div class="carousel-item">
            <img src="static/images/sup_vis_removal_benchmark_00(4).png" class="d-block w-100" alt="Image 4">
          </div>
        </div>
        <button class="carousel-control-prev" type="button" data-bs-target="#removalBenchCarousel" data-bs-slide="prev">
          <span class="carousel-control-prev-icon" aria-hidden="true"></span>
          <span class="visually-hidden">Previous</span>
        </button>
        <button class="carousel-control-next" type="button" data-bs-target="#removalBenchCarousel" data-bs-slide="next">
          <span class="carousel-control-next-icon" aria-hidden="true"></span>
          <span class="visually-hidden">Next</span>
        </button>
      </div>
    </div>
  </section>

  <!-- Citation Section -->
  <section class="py-5 bg-light" style="background: linear-gradient(to bottom, rgb(237, 240, 248), #e9ecef);" data-aos="fade-up" data-aos-delay="400">
    <div class="container">
      <div class="row justify-content-center">
        <div class="col-lg-8">
          <div class="card shadow-sm border-0" data-aos="fade-up">
            <div class="card-header bg-primary text-white">
              <h3 class="mb-0 text-center">Cite Our Work</h3>
            </div>
            <div class="card-body">
              <div class="citation-box p-4 bg-light rounded">
                <pre class="m-0" style="font-family: 'Roboto Mono', monospace; font-size: 0.9rem;">
@article{wei2025omnieraserremoveobjectseffects,
        title={OmniEraser: Remove Objects and Their Effects in Images with Paired Video-Frame Data},
        author={Runpu Wei and Zijin Yin and Shuo Zhang and Lanxiang Zhou and Xueyi Wang and Chao Ban and Tianwei Cao and Hao Sun and Zhongjiang He and Kongming Liang and Zhanyu Ma},
        journal={arXiv preprint arXiv:2501.07397},
        year={2025},
        url={https://arxiv.org/abs/2501.07397},
}</pre>
              </div>
              <div class="text-center mt-4">
                <button class="btn btn-primary mx-2 copy-btn" data-clipboard-target=".citation-box pre">
                  <i class="bi bi-clipboard me-2"></i>Copy Citation
                </button>
                <a href="https://arxiv.org/abs/2501.07397" target="_blank" class="btn btn-outline-primary mx-2">
                  <i class="bi bi-file-earmark-text me-2"></i>View on arXiv
                </a>
              </div>
            </div>
          </div>
        </div>
      </div>
    </div>
  </section>

  <!-- Footer -->
  <footer>
    <div class="py-5 bg-dark text-light text-center">
      <p>This page was built upon the Academic Project Page Template.<br>Licensed under a <a
          href="http://creativecommons.org/licenses/by-sa/4.0/" target="_blank" class="text-light">Creative Commons
          Attribution-ShareAlike 4.0 License</a>.</p>
    </div>
  </footer>

  <!-- Scripts -->
  <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.3.0/dist/js/bootstrap.bundle.min.js"></script>
  <script src="https://unpkg.com/aos@2.3.1/dist/aos.js"></script>
  <script src="https://unpkg.com/swiper/swiper-bundle.min.js"></script>
  <script>
    document.addEventListener('DOMContentLoaded', function () {
      // 初始化轮播图并设置间隔时间（单位：毫秒）
      const videoCarousel = new bootstrap.Carousel(document.getElementById('video4removalCarousel'), {
        interval: 3000  // 3秒切换一次
      });

      const benchCarousel = new bootstrap.Carousel(document.getElementById('removalBenchCarousel'), {
        interval: 3000  // 3秒切换一次
      });
    });
  </script>
  <!-- Clipboard.js -->
  <script src="https://cdnjs.cloudflare.com/ajax/libs/clipboard.js/2.0.8/clipboard.min.js"></script>
  <script>
    // Initialize clipboard
    new ClipboardJS('.copy-btn');

    // Add feedback when copied
    document.querySelectorAll('.copy-btn').forEach(btn => {
      btn.addEventListener('click', function () {
        const originalText = this.innerHTML;
        this.innerHTML = '<i class="bi bi-check-circle me-2"></i>Copied!';
        setTimeout(() => {
          this.innerHTML = originalText;
        }, 2000);
      });
    });
  </script>

  <script>
    document.addEventListener('DOMContentLoaded', function () {
      // Initialize AOS
      AOS.init({
        duration: 1000,
        easing: 'ease-in-out',
      });

      // Preload alternate images
      /*
      const preloadImages = document.querySelectorAll('.alternate');
      preloadImages.forEach((img, index) => {
        if (index < 1) { // 限制同时加载的图片数量，比如 5
          const src = img.src;
          const preloadedImg = new Image();
          preloadedImg.src = src;
        }
      });
      */

    });
  </script>
</body>

</html>