HomePage_D3-Predictor/index.html at main · X-GenGroup/HomePage_D3-Predictor · 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
<!DOCTYPE html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="description"
        content="D³-Predictor: Noise-Free Deterministic Diffusion for Dense Prediction">
  <meta name="keywords" content="From Ideal to Real">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>D³-Predictor: Noise-Free Deterministic Diffusion for Dense Prediction</title>

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

  <link rel="stylesheet" href="./static/css/bulma.min.css">
  <link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
  <link rel="stylesheet" href="./static/css/bulma-slider.min.css">
  <link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
  <link rel="stylesheet"
        href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
  <link rel="stylesheet" href="./static/css/index.css">
  <link rel="icon" href="./static/favicon_io/logo3.png">

  <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
  <script defer src="./static/js/fontawesome.all.min.js"></script>
  <script src="./static/js/bulma-carousel.min.js"></script>
  <script src="./static/js/bulma-slider.min.js"></script>
  <script src="./static/js/index.js"></script>
  <link rel="stylesheet" href="https://cdn.knightlab.com/libs/juxtapose/latest/css/juxtapose.css">
  <script src="https://cdn.knightlab.com/libs/juxtapose/latest/js/juxtapose.min.js"></script>

  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.0/css/all.min.css">
</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">
          <img src="static/favicon_io/logo3.png" alt="Logo" style="height: 100px; margin-bottom: 6px;">
          <h1 class="title is-1 publication-title">D³-Predictor: Noise-Free Deterministic Diffusion for Dense Prediction</h1>
          <div class="is-size-5 publication-authors">
            <span class="author-block">
              <a href="https://xcltql666.github.io/xcll.github.io/">Changliang Xia</a><sup>*</sup>,
            </span>
            <span class="author-block">
              <a href="https://chengyou-jia.github.io/">Chengyou Jia</a><sup>*</sup>,
            </span>
            <span class="author-block">
              <a href="https://scholar.google.com/citations?hl=en&user=C3ujEF0AAAAJ">Minnan Luo</a><sup>†</sup>,
            </span>
            <span class="author-block">
              <a href="https://scholar.google.com/citations?user=8qX4Z5IAAAAJ">Zhuohang Dang</a>,
            </span>
            <span class="author-block">
              <a href="https://github.com/King-Wood-Shen/Xin-Shen.github.io">Xin Shen</a>,
            </span>
            <span class="author-block">
              <a href="https://jayce-ping.github.io/homepage">Bowen Ping</a>,
            </span>
          </div>

          <div class="is-size-5 publication-authors">
            <span class="author-block">Xi'an Jiaotong University</span>
            <br>
            <span class="eql-cntrb"><small><sup>*</sup>Equal Contribution.</small></span>
            <span class="eql-cntrb"><small><sup>†</sup>Corresponding Author.</small></span>
          </div>
          <div class="column has-text-centered">
            <div class="publication-links">
              <!-- PDF Link. -->
              <span class="link-block">
                <a href="https://arxiv.org/pdf/2512.07062"
                   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>
              <span class="link-block">
                <a href="https://arxiv.org/abs/2512.07062"
                   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>

              <span class="link-block">
                <a href="https://github.com/X-GenGroup/D3-Predictor"
                  class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                    <i class="fab fa-github"></i>
                  </span>
                  <span>Code</span>
                </a>
              </span>
              <span class="link-block">
                <a href="https://huggingface.co/collections/X-GenGroup/d3-predictor-model"
                  class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                    <img src="https://huggingface.co/front/assets/huggingface_logo-noborder.svg" alt="Hugging Face" style="height: 1em;">
                  </span>
                  <span>Model</span>
                </a>
              </span>
              <span class="link-block">
                <a href="https://huggingface.co/collections/X-GenGroup/d3-predictor-data"
                  class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                    <img src="https://huggingface.co/front/assets/huggingface_logo-noborder.svg" alt="Hugging Face" style="height: 1em;">
                  </span>
                  <span>Dataset</span>
                </a>
              </span>
            </div>
          </div>
        </div>
      </div>
    </div>
  </div>
</section>

<section class="section hero" style="padding-bottom: 0;">

  <div class="section has-text-centered" style="margin-top: -6em;">
    <div style="display: flex; align-items: center; justify-content: center; gap: 16px; margin-bottom: 32px;">
      <img src="static/images/more.png" alt="Logo" style="height: 28px;">
      <h2 class="title is-3 has-text-centered" style="margin: 0;">Demos and Comparison</h2>
    </div>
    <div class="buttons is-centered" id="demo-buttons">
      <button class="button is-info is-active" onclick="showDemoImage('adverse', 0)">Depth Estimation</button>
      <button class="button is-info" onclick="showDemoImage('city', 1)">Surface Normal Estimation</button>
      <button class="button is-info" onclick="showDemoImage('medical', 2)">Image Matting</button>
      <button class="button is-info" onclick="showDemoImage('eco', 3)">Real-World Applications (1)</button>
      <button class="button is-info" onclick="showDemoImage('safety', 4)">Real-World Applications (2)</button>
    </div>
    <div id="demo-image-container" style="margin-top: 1.5em; position: relative; display: flex; justify-content: center; align-items: center; height: 700px; width: 100%; margin-bottom: 0em; overflow: hidden;">
    </div>
  </div>

  <style>
    #demo-buttons .button.is-active {
      background-color: #23527c !important;
      color: #fff !important;
      box-shadow: 0 4px 16px rgba(35,82,124,0.25), 0 1.5px 4px rgba(0,0,0,0.15);
      border: 2px solid #16324a;
      transition: box-shadow 0.2s, border 0.2s, background 0.2s;
    }
    .demo-image {
      position: absolute;
      top: 50%;
      left: 50%;
      transform: translate(-50%, -50%);
      max-width: 60%;
      max-height: 90%;
      height: auto;
      opacity: 0;
      transition: opacity 0.6s ease-in-out;
    }
    .demo-image.active {
      opacity: 1;
    }
    .demo-image.fade-out {
      opacity: 0;
    }
  </style>

  <script>
  const demoData = [
    {type: 'adverse', img: 'static/images/proj_depth_page-0001.jpg', alt: 'Depth Estimation'},
    {type: 'city', img: 'static/images/proj_normal_page-0001.jpg', alt: 'Surface Normal Estimation'},
    {type: 'medical', img: 'static/images/proj_matting_page-0001.jpg', alt: 'Image Matting'},
    {type: 'eco', img: 'static/images/crop_1_page-0001.jpg', alt: 'Real-World Applications (1)'},
    {type: 'safety', img: 'static/images/crop_2_page-0001.jpg', alt: 'Real-World Applications (2)'}
  ];
  let currentDemoIndex = 0;
  let autoSwitchTimer = null;
  let isSwitching = false; // 标志位，防止切换过程中再次切换

  window.addEventListener('DOMContentLoaded', function() {
    showDemoImage('adverse', 0, true);
  });

  function showDemoImage(type, idx, restartTimer = true) {
    // 如果正在切换中，忽略本次切换请求
    if (isSwitching) {
      return;
    }

    const btns = document.querySelectorAll('#demo-buttons .button');
    btns.forEach((btn, i) => {
      if(i === idx) btn.classList.add('is-active');
      else btn.classList.remove('is-active');
    });

    const demo = demoData.find(d => d.type === type);
    const container = document.getElementById('demo-image-container');
    const oldImg = container.querySelector('.demo-image.active');

    // 如果是同一张图片，不需要切换
    if (oldImg && oldImg.src.endsWith(demo.img)) {
      currentDemoIndex = idx;
      if (restartTimer) {
        scheduleAutoSwitch();
      }
      return;
    }

    // 设置切换标志位
    isSwitching = true;

    // 创建新图片元素
    const newImg = document.createElement('img');
    newImg.className = 'demo-image';
    newImg.src = demo.img;
    newImg.alt = demo.alt;

    // 先加载新图片，加载完成后再开始动画
    newImg.onload = function() {
      // 先将新图片添加到容器（但opacity为0，不可见）
      container.appendChild(newImg);

      // 强制重排，确保新图片已添加到DOM
      void newImg.offsetWidth;

      // 如果有旧图片，先让它淡出
      if (oldImg) {
        oldImg.classList.add('fade-out');
      }

      // 下一帧开始淡入新图片（与旧图片淡出同时进行）
      requestAnimationFrame(() => {
        newImg.classList.add('active');

        // 动画完成后移除旧图片并清除切换标志位
        setTimeout(() => {
          if (oldImg && oldImg.parentNode) {
            oldImg.remove();
          }
          isSwitching = false;
        }, 600); // 与CSS transition时间一致
      });
    };

    // 如果图片已经在缓存中，onload可能不会触发，所以也检查complete
    if (newImg.complete) {
      newImg.onload();
    } else {
      // 如果图片加载失败，至少显示新图片
      newImg.onerror = function() {
        if (oldImg) {
          oldImg.remove();
        }
        container.appendChild(newImg);
        newImg.classList.add('active');
        isSwitching = false;
      };
    }

    currentDemoIndex = idx;
    // 启动/重启自动切换定时器
    if (restartTimer) {
      scheduleAutoSwitch();
    }
  }

  function scheduleAutoSwitch() {
    if (autoSwitchTimer) {
      clearInterval(autoSwitchTimer);
    }
    autoSwitchTimer = setInterval(() => {
      const nextIndex = (currentDemoIndex + 1) % demoData.length;
      const nextDemo = demoData[nextIndex];
      showDemoImage(nextDemo.type, nextIndex, false);
    }, 3000);
  }
  </script>
</section>

<section class="section" style="padding-top: 0;">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div style="display: flex; align-items: center; justify-content: center; gap: 16px; margin-bottom: 32px; margin-top: 0em;">
          <img src="static/images/abstract.png" alt="Logo" style="height: 34px;">
          <h2 class="title is-3">Abstract</h2>
        </div>
        <div class="content has-text-justified">
          <p>
            Although diffusion models with strong visual priors have emerged as powerful dense prediction backboens, they overlook a core limitation: the stochastic noise at the core of diffusion sampling is inherently misaligned with dense prediction that requires a deterministic mapping from image to geometry. In this paper, we show that this stochastic noise corrupts fine-grained spatial cues and pushes the model toward timestep-specific noise objectives, consequently destroying meaningful geometric structure mappings. To address this, we introduce <strong>D³-Predictor</strong>, a noise-free deterministic framework built by reformulating a pretrained diffusion model without stochasticity noise. Instead of relying on noisy inputs to leverage diffusion priors, D³-Predictor views the pretrained diffusion network as an ensemble of timestep-dependent visual experts and self-supervisedly aggregates their heterogeneous priors into a single, clean, and complete geometric prior. Meanwhile, we utilize task-specific supervision to seamlessly adapt this noise-free prior to dense prediction tasks. Extensive experiments on various dense prediction tasks demonstrate that D³-Predictor achieves competitive or state-of-the-art performance in diverse scenarios. In addition, it requires less than half the training data previously used and efficiently performs inference in a single step.
          </p>
        </div>
      </div>
    </div>

    <section class="hero is-small">
      <div class="hero-body">
          <div class="container  is-max-desktop">
              <div class="columns is-centered has-text-centered">
                  <div class="column is-full">
                    <div style="display: flex; align-items: center; justify-content: center; gap: 16px; margin-bottom: 32px;">
                      <img src="static/favicon_io/logo3.png" alt="Logo" style="height: 48px;">
                      <h2 class="title is-3">D³-Predictor Architecture</h2>
                    </div>
                      <div class="item">
                          <img src="static/images/method8.jpg" width="90%" alt="ap-clip" />
                          <h2 class="subtitle has-text-left" style="font-size: 16px;">
                          <strong>Overview of the D³-Predictor.</strong> (a) We reformulate the pretrained diffusion model into a noise-free framework to better suit dense prediction tasks, without compromising the diffusion prior with minimal overhead. (b) The D³-Predictor takes a clean image as input and produces an accurate prediction with impressive geometric details in a single step.
                          </h2>
                      </div>
                  </div>
              </div>
          </div>
      </div>
    </section>

    <section class="section" id="BibTeX">
      <div class="container is-max-desktop content">
        <h2 class="title">BibTeX</h2>
        <p>If you find our paper or code useful for your research, please consider citing our work.</pp>
        <br>
        <pre><code>@misc{xia2025mathrmdmathrm3predictornoisefreedeterministicdiffusion,
          title={D³-Predictor: Noise-Free Deterministic Diffusion for Dense Prediction},
          author={Changliang Xia and Chengyou Jia and Minnan Luo and Zhuohang Dang and Xin Shen and Bowen Ping},
          year={2025},
          eprint={2512.07062},
          archivePrefix={arXiv},
          primaryClass={cs.CV},
          url={https://arxiv.org/abs/2512.07062},
        }</code></pre>
      </div>
    </section>

    <footer class="footer">
      <div class="container">
        <div class="content has-text-centered">
          <a class="icon-link"
             href="" class="external-link" disabled>
            <i class="fas fa-file-pdf"></i>
          </a>
          <a class="icon-link" href="" class="external-link" disabled>
            <i class="fab fa-github"></i>
          </a>
        </div>
        <div class="columns is-centered">
          <div class="column is-8">
            <div class="content">
              <p>
                This website is licensed under a <a rel="license"
                                                    href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
                Commons Attribution-ShareAlike 4.0 International License</a>.
              </p>
              <p>
                This means you are free to borrow the <a
                  href="https://github.com/nerfies/nerfies.github.io">source code</a> of this website,
                we just ask that you link back to this page in the footer.
                Please remember to remove the analytics code included in the header of the website which
                you do not want on your website.
              </p>
            </div>
          </div>
        </div>
      </div>
    </footer>

</body>
</html>