-
Notifications
You must be signed in to change notification settings - Fork 2
Expand file tree
/
Copy pathindex.html
More file actions
346 lines (319 loc) · 11.5 KB
/
Copy pathindex.html
File metadata and controls
346 lines (319 loc) · 11.5 KB
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
<!doctype html>
<html lang="hu">
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width,initial-scale=1" />
<title>SceneVGGT — VGGT-based online 3D semantic SLAM</title>
<meta name="description" content="SceneVGGT: VGGT-based online 3D semantic SLAM for indoor scene understanding and navigation." />
<style>
:root{
--bg:#0b0f19; --card:#0f1629; --text:#e7ecff; --muted:#a9b3d6;
--border:rgba(255,255,255,.10); --accent:#7aa2ff; --accent2:#67e8f9;
--code:#0a1224;
--shadow: 0 12px 30px rgba(0,0,0,.35);
--radius:18px;
color-scheme: dark;
}
*{box-sizing:border-box}
body{
margin:0;
font-family: ui-sans-serif, system-ui, -apple-system, Segoe UI, Roboto, Helvetica, Arial, "Apple Color Emoji","Segoe UI Emoji";
background: radial-gradient(1200px 600px at 10% 0%, rgba(122,162,255,.18), transparent 60%),
radial-gradient(900px 500px at 90% 10%, rgba(103,232,249,.16), transparent 55%),
var(--bg);
color:var(--text);
line-height:1.55;
}
a{color:var(--accent); text-decoration:none}
a:hover{text-decoration:underline}
/* a teljes oldal “konténer” – ez limitálja az egész lap szélességét */
.wrap{max-width:980px; margin:0 auto; padding:40px 18px 64px}
header{
text-align:center;
padding:30px 18px 10px;
}
.badge{
display:inline-flex; gap:8px; align-items:center;
padding:7px 12px;
border:1px solid var(--border);
border-radius:999px;
background: rgba(255,255,255,.03);
color:var(--muted);
font-size:13px;
}
h1{
margin:16px 0 10px;
font-size: clamp(26px, 3.6vw, 40px);
line-height:1.15;
letter-spacing:-.02em;
}
.subtitle{
max-width:820px;
margin: 0 auto 18px;
color:var(--muted);
font-size: 16px;
}
.btns{display:flex; justify-content:center; gap:10px; flex-wrap:wrap; margin:18px 0 10px}
.btn{
display:inline-flex; align-items:center; gap:8px;
padding:10px 14px;
border-radius:12px;
border:1px solid var(--border);
background: rgba(255,255,255,.04);
color:var(--text);
transition: transform .08s ease, background .15s ease, border-color .15s ease;
font-weight:600;
font-size:14px;
}
.btn:hover{background: rgba(255,255,255,.07); border-color: rgba(255,255,255,.16); text-decoration:none}
.btn:active{transform: translateY(1px)}
.btn.primary{
background: linear-gradient(135deg, rgba(122,162,255,.22), rgba(103,232,249,.14));
border-color: rgba(122,162,255,.35);
}
.authors{
margin:14px auto 8px;
color:var(--muted);
font-size:15px;
}
.authors b{color:var(--text); font-weight:700}
.notes{
margin:8px auto 0;
color:var(--muted);
font-size:13px;
}
.grid{
display:grid;
grid-template-columns: 1fr;
gap:16px;
margin-top:22px;
}
@media (min-width: 860px){
.grid{grid-template-columns: 1.15fr .85fr;}
}
/* ez a trükk: elem átnyúlhat mindkét oszlopon */
.span-all{
grid-column: 1 / -1;
}
section.card, aside.card{
background: rgba(255,255,255,.03);
border:1px solid var(--border);
border-radius: var(--radius);
box-shadow: var(--shadow);
padding:18px 18px 16px;
}
section.card h2, aside.card h2{
margin:0 0 10px;
font-size:18px;
letter-spacing:-.01em;
}
p{margin:0 0 12px}
ul{margin:0 0 12px; padding-left:18px; color:var(--muted)}
li{margin:6px 0}
.muted{color:var(--muted)}
/* FULL-WIDTH GIF blokk (a grid teljes szélességén) */
.media{
display:grid;
grid-template-columns: 1fr 1fr;
gap:16px;
width:100%;
margin-top:12px;
}
@media (max-width: 560px){
.media{grid-template-columns:1fr}
}
figure{
margin:0;
border-radius:16px;
overflow:hidden;
border:1px solid var(--border);
background: rgba(0,0,0,.18);
}
figure img{
width:100%;
height:auto;
display:block;
/* ne vágjon le semmit */
object-fit: contain;
}
figcaption{
padding:10px 12px;
color:var(--muted);
font-size:13px;
border-top:1px solid var(--border);
}
pre{
margin:12px 0 0;
padding:14px 14px;
background: rgba(10,18,36,.72);
border:1px solid rgba(255,255,255,.09);
border-radius: 14px;
overflow:auto;
font-size:13px;
line-height:1.45;
}
code{font-family: ui-monospace, SFMono-Regular, Menlo, Monaco, Consolas, "Liberation Mono", "Courier New", monospace;}
.kpi{
display:flex; gap:12px; flex-wrap:wrap; margin-top:10px;
}
.pill{
display:inline-flex; flex-direction:column; gap:2px;
padding:10px 12px;
border-radius:14px;
border:1px solid var(--border);
background: rgba(255,255,255,.03);
min-width: 160px;
}
.pill .t{color:var(--muted); font-size:12px}
.pill .v{font-weight:800; font-size:15px}
footer{
margin-top:22px;
text-align:center;
color:var(--muted);
font-size:13px;
}
.hr{height:1px; background:var(--border); margin:12px 0}
</style>
</head>
<body>
<div class="wrap">
<header>
<div class="badge" title="Project landing page">
<span>🧭</span>
<span>SceneVGGT</span>
<span class="muted">•</span>
<span class="muted">Online 3D semantic SLAM</span>
</div>
<h1>SceneVGGT: VGGT-based online 3D semantic SLAM for indoor scene understanding and navigation</h1>
<p class="subtitle">
<b>SceneVGGT</b> is a spatio-temporal 3D scene understanding framework that combines SLAM with semantic mapping for autonomous and assistive navigation.
It supports online, real-time processing of streamed data (e.g., from an iPhone Pro). The pipeline’s GPU memory usage remains under 17 GB, irrespective
of sequence length, and achieves competitive point-cloud performance on the ScanNet++ benchmark. Overall, SceneVGGT ensures robust semantic
identification and is fast enough to support interactive assistive navigation with audio feedback.
</p>
<div class="btns">
<a class="btn primary" href="https://arxiv.org/abs/2602.15899" target="_blank" rel="noreferrer">📄 Paper (arXiv)</a>
<a class="btn" href="https://github.com/HBVC-AI/SceneVGGT/" target="_blank" rel="noreferrer">💻 Code (GitHub)</a>
<a class="btn" href="#installation">⚙️ Installation</a>
<a class="btn" href="#citation">✍️ Citation</a>
</div>
<div class="authors">
Anna Gelencsér-Horváth<sup>*</sup><sup>†</sup> ·
Gergely Dinya<sup>*</sup> ·
Péter Halász ·
Dorka Erős ·
Islam Muhammad Muqsit ·
Kristóf Karacs
</div>
<div class="notes">
<span><sup>*</sup> Equal contribution.</span>
<span><sup>†</sup> Corresponding author.</span>
</div>
</header>
<div class="grid">
<!-- Bal oldali (szöveg + KPI) -->
<section class="card">
<h2>Overview</h2>
<p class="muted">
SceneVGGT enables temporally coherent 3D semantic mapping by lifting 2D instance masks into 3D and tracking instances with the VGGT tracking head.
Persistent object identities + timestamps provide computationally efficient, temporally consistent change detection, while floor-plane projection of object
locations supports downstream assistive navigation—including a proof-of-concept navigation module.
</p>
<div class="kpi">
<div class="pill">
<div class="t">Processing</div>
<div class="v">Online / real-time</div>
</div>
<div class="pill">
<div class="t">Input</div>
<div class="v">Streamed data (e.g., iPhone Pro)</div>
</div>
<div class="pill">
<div class="t">GPU memory</div>
<div class="v">< 17 GB (sequence-length invariant)</div>
</div>
</div>
</section>
<!-- Jobb oldali (news + download) -->
<aside class="card">
<h2>News</h2>
<ul>
<li><b>[2026/4/30] Paper accepted for the IEEE ICIP 2026 conference.</b> </li>
<li><b>[2026/2/13]</b> Paper released on arXiv.</li>
<li><b>[2025/2/12]</b> Code release.</li>
</ul>
<div class="hr"></div>
<h2>Download checkpoints</h2>
<p class="muted" style="margin-bottom:10px">
Please download VGG-T model checkpoint from Hugging Face:
</p>
<a class="btn" href="https://huggingface.co/facebook/VGGT-1B/blob/main/model.pt" target="_blank" rel="noreferrer">
⬇️ VGGT-1B model.pt
</a>
<div class="hr"></div>
<h2>Evaluation</h2>
<p class="muted">Coming soon.</p>
</aside>
<!-- Itt a lényeg: FULL szélesség a két oszlopon -->
<section class="card span-all">
<h2>3D semantic SLAM and navigation from Streaming Inputs</h2>
<div class="media">
<figure>
<img src="assets/lab.gif" alt="3D semantic SLAM demo (lab.gif)" />
<figcaption>3D semantic SLAM and navigation from streaming inputs</figcaption>
</figure>
<figure>
<img src="assets/change.gif" alt="Change detection demo (change.gif)" />
<figcaption>Temporally consistent change detection</figcaption>
</figure>
</div>
<p style="margin-top: 1rem;">
Additional visualized results on the ScanNet++ dataset are available in this
<a href="https://www.youtube.com/playlist?list=PLVfIQn5pdvRa3ogeO297-4cRHPp03vmfW" target="blank" rel="noopener noreferrer">
YouTube playlist
</a>.
</p>
</section>
</div>
<section id="installation" class="card" style="margin-top:16px">
<h2>Installation</h2>
<ol class="muted" style="margin:0 0 12px; padding-left:18px">
<li style="margin:8px 0">Clone SceneVGGT</li>
</ol>
<pre><code>git clone git@github.com:HBVC-AI/SceneVGGT.git
cd SceneGGT</code></pre>
<ol start="2" class="muted" style="margin:12px 0 12px; padding-left:18px">
<li style="margin:8px 0">Create conda environment</li>
</ol>
<pre><code>conda create -n scenevggt python=3.10
conda activate SceneVGGT</code></pre>
<ol start="3" class="muted" style="margin:12px 0 12px; padding-left:18px">
<li style="margin:8px 0">Install requirements</li>
</ol>
<pre><code>pip install -r requirements.txt</code></pre>
</section>
<section id="citation" class="card" style="margin-top:16px">
<h2>Citation</h2>
<p class="muted">If you find this project helpful, please cite the following paper:</p>
<pre><code>@article{scenevggt,
title={SceneVGGT: VGGT-based Online 3D Semantic SLAM for Indoor Scene Understanding and Navigation},
author={Anna Gelencsér-Horváth, Gergely Dinya, Dorka Boglárka Erős, Péter Halász, Islam Muhammad Muqsit, Kristóf Karacs},
year={2026},
eprint={2602.15899},
archivePrefix={arXiv},
primaryClass={cs.RO},
url={https://arxiv.org/abs/2602.15899},
}</code></pre>
</section>
<footer>
<div class="muted">
© <span id="y"></span> • SceneVGGT • HBVC-AI
</div>
</footer>
</div>
<script>
document.getElementById("y").textContent = new Date().getFullYear();
</script>
</body>
</html>