spacecontrol3d.github.io/index.html at main · spacecontrol3d/spacecontrol3d.github.io · 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
<!DOCTYPE html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="description"
        content="SuperDec">
  <meta name="keywords" content="SpaceControl, Superquadrics, 3D Generation">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>SpaceControl</title>

  <!-- Global site tag (gtag.js) - Google Analytics -->
  <script async src="https://www.googletagmanager.com/gtag/js?id=G-PYVRSFMDRL"></script>
  <script>
    window.dataLayer = window.dataLayer || [];

    function gtag() {
      dataLayer.push(arguments);
    }

    gtag('js', new Date());

    gtag('config', 'G-PYVRSFMDRL');
  </script>

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

  <script src="https://polyfill.io/v3/polyfill.min.js?features=es6"></script>
  <script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>

  <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/images/icon.svg">
  <!-- <link
  rel="stylesheet"
  href="https://cdn.jsdelivr.net/npm/bootstrap@4.3.1/dist/css/bootstrap.min.css"
  integrity="sha384-ggOyR0iXCbMQv3Xipma34MD+dH/1fQ784/j6cY/iJTQUOhcWr7x9JvoRxT2MZw1T"
  crossorigin="anonymous"
  /> -->
  <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>
  <script src="./static/js/video_comparison.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">
          <h1 class="title is-1 publication-title"> <span style="
                background: linear-gradient(90deg, #ff6ec4, #7873f5);
                -webkit-background-clip: text;
                -webkit-text-fill-color: transparent;
                display: inline-block;
                padding-bottom: 0.1em;">
                SpaceControl
              <!-- </span> <img src="static/figures/compressed/logo/chair.png" width="100" style="position: relative; top: 15px; left: -30px;">  -->
            </h1>
          <h2 class="title is-2 publication-title">Introducing Test-Time Spatial Control to 3D Generative Modeling</h2>
          <!-- <h1 class="title is-4" style="color: #5c5c5c;">ICCV 2025 (Oral)</h1> -->

          <div class="is-size-5 publication-authors">
            <span class="author-block">
              <a href="https://elisabettafedele.github.io/">Elisabetta Fedele</a><sup>1,2*</sup>,</span>
            <span class="author-block">
              <a href="https://francisengelmann.github.io/">Francis Engelmann</a><sup>2*</sup>
            </span>
            <span class="author-block">
              <a href="https://ianhuang.ai">Ian Huang</a><sup>2</sup>,</span>
            <span class="author-block">
            <span class="author-block">
              <a href="https://orlitany.github.io">Or Litany</a><sup>3</sup>,</span>
            <span class="author-block">
              <a href="https://people.inf.ethz.ch/pomarc">Marc Pollefeys</a><sup>1</sup>,
            </span>
            <span class="author-block">
              <a href="https://geometry.stanford.edu/?member=guibas">Leonidas Guibas</a><sup>2</sup>
            </span>

          </div>

          <div class="is-size-5 publication-authors">
            <span class="author-block"><sup>1</sup>ETH Zurich</span> &nbsp;&nbsp;&nbsp;&nbsp;
            <span class="author-block"><sup>2</sup>Stanford University</span> &nbsp;&nbsp;&nbsp;&nbsp;
            <span class="author-block"><sup>3</sup>Technion</span> &nbsp;&nbsp;&nbsp;&nbsp;
          </div>

          <div class="column has-text-centered">
            <div class="publication-links">

              <!-- Code Link. -->

              <span class="link-block">
                <a href="https://arxiv.org/abs/2512.05343"
                   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://github.com/spacecontrol3d/spacecontrol"
                   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="#videos" class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="fab fa-youtube"></i>
                  </span>
                  <span>Videos</span>
                  </a>
              </span>

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

  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-full-width">
        <div class="content has-text-justified">
          <img src="images/teaser_2.jpg">
          <p><strong>Overview of SpaceControl</strong>. SpaceControl enables spatially controlled 3D asset generation using simple geometric primitives such as superquadrics (light blue) or other geometry (e.g., meshes). Top: rapid asset generation. From quick 3D sketches and brief text prompts, we can generate high quality assets. Bottom: fine-grained editing, including adjusting a chair's backrest and adding armrests (left) or precisely controlling a sofa’s dimensions and pillow arrangements (right).</p>
        </div>
      </div>
    </div>
  </div>


<section class="section">
  <div class="container is-max-desktop">
    <!-- Abstract. -->
    <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">
          Generative methods for 3D assets have recently achieved remarkable progress,
        yet providing intuitive and precise control over the object geometry remains a key challenge.
        Existing approaches predominantly rely on text or image prompts, which often fall short in geometric specificity:
        language can be ambiguous, and images are cumbersome to edit. In this work, we introduce SpaceControl, a training-free
        test-time method for explicit spatial control of 3D generation. Our approach accepts diverse geometric inputs,
        from coarse primitives to detailed meshes, and conditions a powerful pre-trained generative model without additional training.
        A controllable parameter lets users trade off between geometric fidelity and output realism. Extensive quantitative evaluation
        and user studies demonstrate that SpaceControl outperforms both training-based and optimization-based baselines in geometric faithfulness
        while preserving high visual quality. Finally, we present an interactive user interface that enables online editing
        of superquadrics for direct conversion into textured 3D assets, facilitating practical deployment in creative workflows.
      </div>
      </div>
    </div>

  </div>
</section>


<section class="section" style="width: 85%; margin: 0 auto">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-full-width">
        <h2 class="title is-3">Method</h2>
        <div class="content has-text-justified">
        <p>
  Given an input conditioning which includes a spatial control, a text prompt and an image (optional),
  <strong>SpaceControl</strong> produces realistic 3D assets.
  First the different conditionings are encoded in a latent space.
  Specifically, the spatial control is voxelized and encoded by Trellis's encoder, the text is encoded by a CLIP encoder,
  and the image (if present) is encoded by a DINOv2 encoder.
  The obtained latents <span>\( z_{0,c} \)</span> are noised up to <span>\( t_0 \)</span> to obtain <span>\( z_{t_0} \)</span>.
  From <span>\( t_0 \)</span> to <span>\( t = 0 \)</span>, <span>\( z_{t_0} \)</span> are denoised by the Structure Flow Model (FM),
  guided by the text prompt features.
  The clean latents <span>\( z_0 \)</span> are then fed into the decoder <span>\( D \)</span>,
  which outputs the voxel grid <span>\( x_0 \)</span>.
  Then, the active voxels are augmented with point-wise noisy latent features,
  denoised by the Appearance Flow Model (FM), using either text or image conditioning.
  The clean latents can then be decoded into versatile output formats such as 3D gaussians (GS),
  radiance fields (RF), and meshes (M) via specific decoders
  <span>\( \mathcal{D}_{O} = \{ \mathcal{D}_{\mathrm{GS}}, \mathcal{D}_{\mathrm{RF}}, \mathcal{D}_{\mathrm{M}} \} \)</span>.
</p>
</div>

        <div class="columns is-1 is-multiline is-mobile">

          <div class="content has-text-justified">
          <figure id="multi-control" style="text-align: center; margin-top: 1.5rem;">
          <img
            src="images/method.png"
            alt="Multimodal Control"
            style="width: 90%; object-fit: cover;"
          />

          </figure>
        </div>

        </div>
        <br>
        <h2 class="title is-3">Controlled 3D Object Generation</h2>
        <h2
          class="title is-4"
          style="text-align: is-justify-content-center; padding-bottom: 10px" id="videos"
        >
        Example 1
        </h2>
        <div class="columns is-1 is-multiline is-mobile">

          <!-- Video Column -->
          <div class="column is-flex is-justify-content-center">
            <video id="robots" autoplay muted loop playsinline controls style="width: 700px; height: auto;">
              <source src="videos/clip_1.mp4" type="video/mp4">
            </video>
          </div>

        </div>
        <h2
          class="title is-4"
          style="text-align: is-justify-content-center; padding-bottom: 10px"
        >
        Example 2
        </h2>
        <div class="columns is-1 is-multiline is-mobile">


          <!-- Video Column -->
          <div class="column is-flex is-justify-content-center">
            <video id="robots" autoplay muted loop playsinline controls style="width: 700px; height: auto;">
              <source src="videos/clip_2.mp4" type="video/mp4">
            </video>
          </div>
        </div>

        <div class="column is-full-width ">
        <h2 class="title is-3">Prompt Adherence Control Strength</h2>
        <div class="content has-text-justified">
      <p>SpaceControl enables different control strengths which lead to varying levels of adherence to the input geometric prompts. By adjusting the control strength, users can influence the balance between geometric fidelity and visual realism in the generated 3D assets.</p>
        <br></div>

        <div class="columns is-1 is-multiline is-mobile">


          <!-- Video Column -->
          <div class="column is-flex is-justify-content-center">
            <video id="robots" autoplay muted loop playsinline controls style="width: 700px; height: auto;">
              <source src="videos/elephants.mp4" type="video/mp4">
            </video>
          </div>
        </div>


    </div>

    <div class="column is-full-width">
        <h2 class="title is-3">Multi-Modal Control</h2>
        <div class="content has-text-justified">
          <p>SpaceControl supports multi-modal control for 3D asset generation by combining spatial guidance via superquadrics with natural language and optional image conditioning. While the model can synthesize assets using only superquadrics and textual prompts, images are particularly useful for maintaining visual consistency during object edits.</p>
          <br>
        </div>

        <div class="columns is-1 is-multiline is-mobile">
          <!-- Video Column -->
          <figure id="multi-control" style="text-align: center; margin-top: 1.5rem;">
          <img
            src="images/text_vs_image.jpg"
            alt="Multimodal Control"
            style="width: 40%; object-fit: cover;"
          />
          <figcaption style="margin-top: 0.75rem;">
            <strong>Multimodal Control.</strong> In addition to geometric conditioning, SpaceControl enables control with text and images.
          </figcaption>
          </figure>
        </div>

    </div>
  </div>
</div>
      <!-- <br>
      <br> -->
    <!-- <div class="container is-max-desktop">
      <h2
          class="title is-4"
          style="text-align: is-justify-content-center; padding-bottom: 10px"
        >
        Qualitative Examples
        </h2>
      <div class="content has-text-justified">
        <br></div>
      <figure id="tsne-embs" style="text-align: center; margin-top: 1.5rem;">
            <img
              src="images/generative.jpg"
              alt="t-SNE visualization of primitive embeddings"
              style="width: 40%; object-fit: cover;"
            />
            <figcaption style="margin-top: 0.75rem;">
              <strong>t-SNE Visualization of Primitive Embeddings</strong> across different ShapeNet classes.
            </figcaption>
          </figure>


  </div> -->
</section>

<section class="section" id="BibTeX">
  <div class="container is-max-desktop content">
    <h2 class="title">BibTeX</h2>
    <pre><code>@article{fedele2025spacecontrol,
   title = {{SpaceControl: Introducing Test-Time Spatial Control to 3D Generative Modeling}},
   author = {Fedele, Elisabetta and Engelmann, Francis and Huang, Ian and Litany, Or and Pollefeys, Marc and Guibas, Leonidas},
   journal = {arXiv preprint arXiv:2512.05343},
   year = {2025}
 }</code></pre>
  </div>
</section>


<footer class="footer">
  <div class="container">
    <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>
            Website source code based on the <a rel="nerfies" href="https://nerfies.github.io/">Nerfies</a> project page
          </p>
        </div>
      </div>
    </div>
  </div>
</footer>

<script src="http://code.jquery.com/jquery.js"></script>

<!-- Fancybox -->
<script src="fancybox/fancybox.umd.js"></script>
<script>
  Fancybox.bind("[data-fancybox]", {
    infinite: false,
    Thumbs: {
      autoStart: false,
    },
  });
</script>

<!-- Javascript Functions -->
<script type="text/javascript">
  let slideIndex = [1, 1, 1, 1, 1];
  let slideId = 0;
  showSlides(slideId, slideIndex[slideId]);
  showSlides(1, slideIndex[1]);
  showSlides(2, slideIndex[2]);
  showSlides(3, slideIndex[3]);
  showSlides(4, slideIndex[4]);

  // Next/previous controls
  function plusSlides(id, n) {
    showSlides(id, (slideIndex[id] += n));
  }

  // Thumbnail image controls
  function currentSlide(id, n) {
    showSlides(id, (slideIndex[id] = n));
  }

  function showSlides(id, n) {
    let i;
    let slides = document.getElementsByClassName("mySlides" + id);
    let dots = document.getElementsByClassName("dot" + id);
    if (n > slides.length) {
      slideIndex[id] = 1;
    }
    if (n < 1) {
      slideIndex[id] = slides.length;
    }
    for (i = 0; i < slides.length; i++) {
      slides[i].style.display = "none";
    }
    for (i = 0; i < dots.length; i++) {
      dots[i].className = dots[i].className.replace(" active", "");
    }
    slides[slideIndex[id] - 1].style.display = "block";
    dots[slideIndex[id] - 1].className += " active";
  }

  // Automatic slideshow
  setInterval(() => {
    plusSlides(0, 1);
  }, 150000); // Change image every 5 seconds
</script>

</body>
</html>