gradientspaces.github.io/index.html at main · GradientSpaces/gradientspaces.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
<!doctype html>
<html lang="en">
  <head>
    <meta charset="utf-8">
    <meta name="viewport" content="width=device-width, initial-scale=1">
    <meta
      name="description"
      content="Gradient Spaces Lab at Stanford builds data-driven methods for spatial intelligence across physical and digital environments."
    >
    <title>Gradient Spaces Lab | Stanford</title>
    <meta property="og:title" content="Gradient Spaces Lab | Stanford">
    <meta property="og:description" content="Data-driven methods for spatial intelligence across physical and digital environments.">
    <meta property="og:type" content="website">
    <meta property="og:url" content="https://gradientspaces.github.io/">
    <meta property="og:image" content="https://gradientspaces.github.io/assets/images/group-photo.jpg">
    <link rel="canonical" href="https://gradientspaces.github.io/">
    <link rel="icon" type="image/svg+xml" href="assets/favicon.svg">
    <link rel="stylesheet" href="styles.css">
  </head>
  <body>
    <header class="site-header">
      <a class="brand" href="index.html">Gradient Spaces</a>
      <button class="nav-toggle" type="button" aria-expanded="false" aria-controls="site-nav" title="Open navigation">
        <span class="sr-only">Menu</span>
        <span aria-hidden="true"></span>
        <span aria-hidden="true"></span>
        <span aria-hidden="true"></span>
      </button>
      <nav class="site-nav" id="site-nav" aria-label="Main navigation">
        <a class="active" href="index.html">Home</a>
        <a href="people.html">People</a>
        <a href="research.html">Research</a>
        <a href="#news">News</a>
        <a href="#connect">Connect</a>
      </nav>
    </header>

    <main>
      <section class="hero">
        <div class="hero-copy">
          <p class="eyebrow">Stanford Civil &amp; Environmental Engineering</p>
          <h1>
            <span>Gradient</span>
            <span>Spaces</span>
            <span>Lab</span>
          </h1>
          <p class="lead">
            We study how real-world visual data can help design, construct, and understand adaptive spaces
            that move between the physical and digital world.
          </p>
          <div class="hero-actions" aria-label="Primary links">
            <a class="strong-link" href="research.html">Research</a>
            <a class="strong-link" href="people.html">People</a>
            <a class="strong-link" href="https://github.com/GradientSpaces">Github</a>
          </div>
        </div>

        <figure class="hero-media">
          <img
            src="assets/images/group-photo.jpg"
            alt="Gradient Spaces group photo at Stanford"
          >
          <figcaption>Gradient Spaces Lab at Stanford.</figcaption>
        </figure>
      </section>

      <section class="section split-intro" aria-labelledby="about-title">
        <div>
          <p class="section-kicker">About</p>
          <h2 id="about-title">Learning spatial futures from the world as it is.</h2>
        </div>
        <div class="body-copy">
          <p>
            Gradient Spaces develops quantitative methods for environments that blend real reality,
            mixed reality, and virtual reality. The lab works across visual data, 3D understanding,
            design, construction, and human experience.
          </p>
          <p>
            Our research spans dynamic 3D scene understanding, in-the-wild reconstruction, mixed
            reality, and sustainable built environments — with open datasets, code, and benchmarks
            released alongside our publications.
          </p>
        </div>
      </section>

      <section class="section" id="news" aria-labelledby="news-title">
        <div class="section-heading">
          <p class="section-kicker">News</p>
          <h2 id="news-title">Latest signals</h2>
        </div>
        <div class="news-list">
          <article class="news-item">
            <time datetime="2026">2026</time>
            <div>
              <h3>Three papers accepted to CVPR 2026</h3>
              <p>Selected work spans dynamic indoor scenes, robust pose estimation, and in-the-wild 3D reconstruction.</p>
            </div>
          </article>
          <article class="news-item">
            <time datetime="2025">2025</time>
            <div>
              <h3>U. V. Helava Award for Best Paper</h3>
              <p>Tao Sun, Iro Armeni, and collaborators were recognized for work on evolving 3D scenes.</p>
            </div>
          </article>
          <article class="news-item">
            <time datetime="2025">2025</time>
            <div>
              <h3>Two papers accepted to NeurIPS 2025</h3>
              <p>GuideFlow3D and Rectified Point Flow join the lab's growing archive of spatial AI work.</p>
            </div>
          </article>
        </div>
      </section>

      <section class="section" aria-labelledby="featured-title">
        <div class="section-heading inline-heading">
          <div>
            <p class="section-kicker">Research</p>
            <h2 id="featured-title">Featured projects</h2>
          </div>
          <a class="text-link" href="research.html">View all research</a>
        </div>

        <div class="project-grid">
          <article class="project-card">
            <img
              src="assets/images/projects/rescene4d.jpeg"
              alt="ReScene4D project teaser"
              loading="lazy"
            >
            <div>
              <p class="tag">CVPR 2026</p>
              <h3>ReScene4D</h3>
              <p>Temporally consistent semantic instance segmentation for evolving indoor 3D scenes.</p>
            </div>
          </article>
          <article class="project-card">
            <img
              src="assets/images/projects/wildpose.jpg"
              alt="WildPose project teaser"
              loading="lazy"
            >
            <div>
              <p class="tag">CVPR 2026</p>
              <h3>WildPose</h3>
              <p>A unified framework for robust pose estimation in unconstrained visual settings.</p>
            </div>
          </article>
          <article class="project-card">
            <img
              src="assets/images/projects/gaussfusion.png"
              alt="GaussFusion project teaser"
              loading="lazy"
            >
            <div>
              <p class="tag">CVPR 2026</p>
              <h3>GaussFusion</h3>
              <p>Geometry-informed video generation for stronger 3D reconstruction in the wild.</p>
            </div>
          </article>
        </div>
      </section>

      <section class="section connect-band" id="connect" aria-labelledby="connect-title">
        <div>
          <p class="section-kicker">Connect</p>
          <h2 id="connect-title">Build, visit, collaborate.</h2>
        </div>
        <div class="link-columns">
          <a href="https://github.com/GradientSpaces">Github</a>
          <a href="https://huggingface.co/gradient-spaces">HuggingFace</a>
          <a href="https://www.linkedin.com/company/gradientspaces">LinkedIn</a>
          <a href="https://cee.stanford.edu/academics-admission/graduate-admissions">PhD in CEE</a>
          <a href="mailto:iroarmeni@stanford.edu">Email</a>
        </div>
      </section>
    </main>

    <footer class="site-footer">
      <p>Gradient Spaces Research Group</p>
      <p>Y2E2, 473 Via Ortega, Room 233, Stanford, CA 94305</p>
    </footer>

    <script src="script.js"></script>
  </body>
</html>