-
Notifications
You must be signed in to change notification settings - Fork 3
Expand file tree
/
Copy pathchallenge.html
More file actions
583 lines (547 loc) · 27.1 KB
/
challenge.html
File metadata and controls
583 lines (547 loc) · 27.1 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
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
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
<!DOCTYPE html>
<html>
<head>
<meta charset="UTF-8">
<title>EmbodiedBench Challenge @ CVPR 2026</title>
<!-- Favicon -->
<link rel="icon" href="website/img/embodied-logo.png" type="image/png">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">
<link rel="stylesheet" href="website/css/bulma.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="website/css/fontawesome.all.min.css">
<script defer src="website/javascript/fontawesome.all.min.js"></script>
<link rel="stylesheet" href="website/css/index.css">
<style>
.challenge-hero {
background: linear-gradient(135deg, #f8f9fa 0%, #e9ecef 100%);
padding: 2rem 0;
}
.challenge-card {
border: 1px solid #e0e0e0;
border-radius: 8px;
padding: 1.5rem;
margin-bottom: 1rem;
background: #fff;
}
.challenge-card h4 {
margin-bottom: 0.5rem;
color: #363636;
}
.data-split-card {
display: flex;
align-items: center;
padding: 1rem 1.5rem;
border-radius: 8px;
margin-bottom: 1rem;
background: #f8f9fa;
border: 1px solid #e9ecef;
}
.data-split-icon {
font-size: 2rem;
margin-right: 1rem;
}
.data-split-content {
flex: 1;
}
.data-split-badge {
padding: 0.25rem 0.75rem;
border-radius: 4px;
font-size: 0.85rem;
font-weight: 600;
}
.badge-public { background: #d4edda; color: #155724; }
.badge-coming { background: #fff3cd; color: #856404; }
.nav-link-challenge {
display: inline-flex;
align-items: center;
padding: 0.5rem 1rem;
margin: 0.25rem;
background: #5e7154;
color: white !important;
border-radius: 6px;
text-decoration: none;
font-weight: 500;
transition: background 0.2s;
}
.nav-link-challenge:hover {
background: #4a5d42;
color: white !important;
}
.leaderboard-table {
width: 100%;
border-collapse: collapse;
margin-top: 1rem;
}
.leaderboard-table th,
.leaderboard-table td {
padding: 0.75rem 1rem;
text-align: left;
border-bottom: 1px solid #e0e0e0;
}
.leaderboard-table th {
background: #f8f9fa;
font-weight: 600;
cursor: pointer;
}
.leaderboard-table th:hover {
background: #e9ecef;
}
.code-block {
background: #2d2d2d;
color: #f8f8f2;
padding: 1rem;
border-radius: 6px;
overflow-x: auto;
white-space: pre;
font-family: 'Monaco', 'Menlo', monospace;
font-size: 0.9rem;
line-height: 1.5;
}
.rules-list {
list-style: none;
padding-left: 0;
}
.rules-list li {
padding: 0.5rem 0;
border-bottom: 1px solid #f0f0f0;
display: flex;
align-items: flex-start;
}
.rules-list li:last-child {
border-bottom: none;
}
.rules-list .rule-icon {
margin-right: 0.75rem;
color: #5e7154;
}
.timeline-item {
border-left: 3px solid #5e7154;
padding: 0.75rem 1.25rem;
margin-bottom: 1rem;
background: #f8f9fa;
border-radius: 0 8px 8px 0;
}
.timeline-item .timeline-label {
font-weight: 700;
color: #5e7154;
font-size: 1rem;
}
.timeline-item .timeline-date {
font-size: 0.9rem;
color: #555;
margin-bottom: 0.25rem;
}
</style>
</head>
<body>
<!-- Navigation Bar -->
<nav class="navbar" role="navigation" aria-label="main navigation" style="background: #fff; box-shadow: 0 1px 3px rgba(0,0,0,0.1);">
<div class="container">
<div class="navbar-brand">
<a class="navbar-item" href="index.html">
<img src="website/img/embodied-logo.png" alt="EmbodiedBench" width="40" height="40">
<span class="embodiedbench" style="margin-left: 0.5rem; font-size: 1.25rem;">EmbodiedBench</span>
</a>
</div>
<div class="navbar-end">
<a href="index.html" class="navbar-item">Home</a>
<a href="challenge.html" class="navbar-item has-text-weight-bold" style="color: #5e7154;">Challenge</a>
</div>
</div>
</nav>
<!-- Hero Section -->
<section class="challenge-hero">
<div class="container">
<div class="has-text-centered">
<h1 class="title is-2" style="margin-bottom: 0.5rem;">
<img src="website/img/embodied-logo.png" alt="logo" width="60" height="60" style="vertical-align: middle; margin-right: 10px;">
<span class="embodiedbench">EmbodiedBench</span> Challenge
</h1>
<p class="subtitle is-4" style="color: #4a4a4a; margin-top: 0.5rem;">
Benchmarking Vision-Driven Embodied Agents across EB-ALFRED and EB-Navigation
</p>
<p class="is-size-5" style="margin-top: 1rem;">
🏛️ <strong>Hosted by:</strong> <a href="https://foundation-models-meet-embodied-agents.github.io/cvpr2026/" target="_blank">Foundation Models Meet Embodied Agents Workshop</a>, CVPR 2026.
</p>
<p class="content" style="max-width: 800px; margin: 1rem auto;">
The <strong>EmbodiedBench Challenge</strong> evaluates Multi-modal Large Language Models (MLLMs) as vision-driven embodied agents across two environments: <strong>EB-ALFRED</strong> (household task planning) and <strong>EB-Navigation</strong> (spatial navigation). <strong>Stage 1 runs from April 15, 2026 to May 25, 2026, 23:59 AoE (Anywhere on Earth)</strong>, and participants can submit trajectory results via EvalAI. The top 5 teams will advance to Stage 2 for held-out evaluation, with full Stage 2 instructions available below.
</p>
<!-- Quick Links -->
<div class="buttons is-centered" style="margin-top: 1.5rem; flex-wrap: wrap;">
<a href="#overview" class="button is-dark is-rounded">Challenge Instructions</a>
<a href="#stage2" class="button is-dark is-rounded">Stage 2 Details</a>
<a href="#awards" class="button is-dark is-rounded">Awards</a>
<a href="#contact" class="button is-dark is-rounded">
<span class="icon"><i class="fas fa-address-book"></i></span>
<span>Contact</span>
</a>
<a href="https://github.com/EmbodiedBench/EmbodiedBench" class="button is-dark is-rounded" target="_blank">
<span class="icon"><i class="fab fa-github"></i></span>
<span>Dataset & Code</span>
</a>
<a href="https://eval.ai/web/challenges/challenge-page/2670/overview" class="button is-dark is-rounded" target="_blank">
<span>Submit on EvalAI</span>
</a>
<a href="https://huggingface.co/EmbodiedBench" class="button is-dark is-rounded" target="_blank">
<span class="icon"><i class="fas fa-database"></i></span>
<span>Dataset</span>
</a>
</div>
</div>
</div>
</section>
<!-- Timeline -->
<section class="section" id="timeline">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Timeline</h2>
<p class="content has-text-centered" style="color: #666; margin-bottom: 1.5rem;">All deadlines use AoE (Anywhere on Earth, UTC−12).</p>
<div class="timeline-item">
<div class="timeline-label">Stage 1 — Qualification Phase</div>
<div class="timeline-date">April 15, 2026 to May 25, 2026, 23:59 AoE (Anywhere on Earth)</div>
<p>Submit trajectory results via EvalAI. Up to 5 submissions per day, 50 total. Leaderboard is private (visible to host only).</p>
</div>
<div class="timeline-item">
<div class="timeline-label">Stage 2 — Final Held-out Test Phase</div>
<div class="timeline-date">May 25, 2026 to May 28, 2026, 23:59 AoE (Anywhere on Earth)</div>
<p>This stage is described publicly below. The top 5 teams from Stage 1 will receive a private submission link, and each team submits <strong>exactly once</strong> by <strong>May 26, 2026 (AoE)</strong>. The remaining Stage 2 period is reserved for online evaluation and addressing engineering issues related to customization. Teams may submit either a vLLM-servable model or a modified EmbodiedBench agent framework.</p>
</div>
<div class="timeline-item">
<div class="timeline-label">Results Announcement</div>
<div class="timeline-date">May 31, 2026</div>
<p>Final rankings and award ceremony at the Foundation Models Meet Embodied Agents Workshop.</p>
</div>
</div>
</section>
<!-- Awards -->
<section class="section" id="awards">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Awards & Recognition</h2>
<div class="challenge-card">
<p class="content">
Top-performing teams will be recognized with both cash awards and workshop visibility opportunities:
</p>
<ul class="content">
<li><strong>1st place:</strong> $500</li>
<li><strong>2nd place:</strong> $300</li>
<li><strong>3rd place:</strong> $200</li>
<li><strong>Certificates:</strong> Award certificates will be provided.</li>
<li><strong>Workshop opportunities:</strong> Selected teams may be invited to submit a technique report and give an on-site talk, subject to the final workshop schedule.</li>
</ul>
</div>
</div>
</section>
<!-- Challenge Overview -->
<section class="section" id="overview">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Challenge Overview</h2>
<div class="challenge-card">
<h4 class="title is-5">Goal</h4>
<p class="content">
Given visual observations and language instructions, develop vision-driven agents that can plan and execute tasks in two EmbodiedBench environments:
</p>
<ul class="content">
<li><strong>EB-ALFRED</strong> — Household task planning: multi-step high-level tasks requiring object interaction, task decomposition, and long-horizon planning.</li>
<li><strong>EB-Navigation</strong> — Spatial navigation: low-level action planning requiring spatial reasoning, visual perception, and precise movement control.</li>
</ul>
</div>
<div class="challenge-card">
<h4 class="title is-5">Stage 1 — What You Do</h4>
<ol class="content" style="padding-left: 1.5rem;">
<li>Run your model or agent using the official EmbodiedBench code on EB-ALFRED and EB-Navigation.</li>
<li>Use <code>copy_json.py</code> to extract episode result files and prepare a zip submission.</li>
<li>Submit the zip file to EvalAI. The evaluator automatically computes your scores and updates the leaderboard.</li>
</ol>
</div>
<div class="challenge-card">
<h4 class="title is-5">Stage 2 — What You Do (Top 5 Teams Only)</h4>
<ol class="content" style="padding-left: 1.5rem;">
<li>Receive a private invitation and submission link from the organizers.</li>
<li>Submit your model, modified agent framework, or a combined model-plus-agent submission (exactly one submission) following one of the three supported formats (see <a href="#stage2">Stage 2 Submission</a> below).</li>
<li>The organizers run your system on held-out EB-ALFRED and EB-Navigation tasks to determine the final ranking.</li>
<li>These held-out tasks will be released after the end of Stage 2.</li>
</ol>
</div>
</div>
</section>
<!-- Evaluation -->
<section class="section" id="evaluation">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Evaluation</h2>
<div class="content">
<p><strong>Leaderboard Metrics:</strong></p>
<ul>
<li><strong>Overall Score</strong> (primary): Average task success rate across EB-ALFRED and EB-Navigation.</li>
<li><strong>ALFRED SR</strong>: Task success rate on EB-ALFRED (%).</li>
<li><strong>ALFRED Steps</strong>: Average steps per episode on EB-ALFRED (lower is better).</li>
<li><strong>Navigation SR</strong>: Task success rate on EB-Navigation (%).</li>
<li><strong>Navigation Steps</strong>: Average steps per episode on EB-Navigation (lower is better).</li>
<li><strong>Average Steps</strong>: Average of per-environment avg steps (tiebreaker, lower is better).</li>
</ul>
<p><strong>Ranking:</strong> Teams are ranked by Overall Score. Ties are broken by Average Steps (fewer is better).</p>
<p><strong>Model size:</strong> Open-source base models must have fewer than <strong>10B parameters</strong>. Commercial APIs are not allowed.</p>
</div>
</div>
</section>
<!-- Leaderboard -->
<section class="section" id="leaderboard">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Challenge Leaderboard</h2>
<p class="content has-text-centered" style="margin-bottom: 1.5rem;">
Stage 1 results on the validation set. Click on column headers to sort.
</p>
<div class="box">
<table class="leaderboard-table">
<thead>
<tr>
<th>Rank ↕</th>
<th>Team / Method ↕</th>
<th>Overall Score ↕</th>
<th>ALFRED SR ↕</th>
<th>ALFRED Steps ↕</th>
<th>Navigation SR ↕</th>
<th>Navigation Steps ↕</th>
<th>Avg Steps ↕</th>
</tr>
</thead>
<tbody>
<tr>
<td colspan="8" class="has-text-centered" style="color: #888; padding: 2rem;">
Stage 1 runs from April 15, 2026 to May 25, 2026, 23:59 AoE (Anywhere on Earth). Leaderboard entries will appear here as submissions are processed.
</td>
</tr>
</tbody>
</table>
</div>
<p class="content has-text-centered" style="font-size: 0.9rem; color: #666;">
The live leaderboard is available on <a href="https://eval.ai/web/challenges/challenge-page/2670/overview" target="_blank">the EvalAI challenge page</a> during Stage 1.
</p>
</div>
</section>
<!-- Stage 1 Submission -->
<section class="section" id="submission">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Stage 1 Submission</h2>
<div class="challenge-card">
<h4 class="title is-5">Step 1 — Run EmbodiedBench</h4>
<p class="content">
Run your model or agent using the official <a href="https://github.com/EmbodiedBench/EmbodiedBench" target="_blank">EmbodiedBench code</a> to generate rollout trajectories for EB-ALFRED and EB-Navigation. A typical raw rollout directory looks like:
</p>
<div class="code-block">eb_alfred/model_experiment/
├── summary_all.json
├── base/
│ ├── episode_1_step_19.json
│ ├── episode_2_step_10.json
│ ├── images/
│ └── results/
│ ├── episode_1_final_res.json
│ └── ...
└── common_sense/
├── episode_1_step_14.json
├── images/
└── results/
└── ...</div>
</div>
<div class="challenge-card">
<h4 class="title is-5">Step 2 — Extract Episode JSON Files</h4>
<p class="content">
Use <a href="https://github.com/EmbodiedBench/EmbodiedBench/blob/master/copy_json.py" target="_blank"><code>copy_json.py</code></a> to extract only the episode JSON files. Run the script <strong>once per environment</strong>:
</p>
<div class="code-block">python copy_json.py --source path/to/running/eb_alfred/model_experiment \
--output submission_folder/eb_alfred
python copy_json.py --source path/to/running/eb_nav/model_experiment \
--output submission_folder/eb_nav</div>
<p class="content" style="margin-top: 1rem;">This produces a compact directory:</p>
<div class="code-block">submission_folder/
├── eb_alfred/
│ └── model_experiment/
│ ├── base/
│ │ ├── episode_1_step_19.json
│ │ └── ...
│ └── common_sense/
│ └── ...
└── eb_nav/
└── model_experiment/
├── base/
│ ├── episode_1.json
│ └── ...
└── common_sense/
└── ...</div>
</div>
<div class="challenge-card">
<h4 class="title is-5">Step 3 — Compress and Submit</h4>
<div class="code-block">cd submission_folder && zip -r ../submission.zip . && cd ..</div>
<p class="content" style="margin-top: 1rem;">
Upload <code>submission.zip</code> to <a href="https://eval.ai/web/challenges/challenge-page/2670/overview" target="_blank">the EvalAI challenge page</a>. The evaluator will automatically extract the zip, compute your scores, and update the leaderboard.
</p>
<p class="content">
<strong>Submission limits:</strong> Up to 5 submissions per day, 50 total during Stage 1.<br>
<strong>Stage 1 timeline:</strong> April 15, 2026 to May 25, 2026, 23:59 AoE (Anywhere on Earth).
</p>
</div>
</div>
</section>
<!-- Stage 2 Submission -->
<section class="section" id="stage2">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Stage 2 Submission (Top 5 Teams)</h2>
<p class="content">
The top 5 teams from Stage 1 will be invited to Stage 2 via a <strong>private link provided by the organizers</strong>. Each team may submit <strong>exactly once</strong>. The submission must be a runnable model, modified agent framework, or combined model-plus-agent system for EmbodiedBench. Three submission formats are supported:
</p>
<p class="content">
<strong>Note:</strong> The deadline for submitting the Stage 2 model is <strong>May 26, 2026 (AoE)</strong>, to allow sufficient time for online evaluation and to address any engineering issues related to customization from participating teams.
<br><strong>Held-out tasks:</strong> Stage 2 uses held-out tasks from EB-ALFRED and EB-Navigation, which will be released after Stage 2 concludes.
</p>
<div class="challenge-card">
<h4 class="title is-5">Option A — vLLM-Compatible Model Server</h4>
<p class="content">
Submit a fine-tuned or adapted model that can be served with <a href="https://docs.vllm.ai" target="_blank">vLLM</a>. The organizers will launch it as:
</p>
<div class="code-block">vllm serve <your-model> --host 0.0.0.0 --port 8000</div>
<p class="content" style="margin-top: 1rem;">
Provide the model path or Hugging Face model ID and any required <code>vllm serve</code> flags.
</p>
</div>
<div class="challenge-card">
<h4 class="title is-5">Option B — Modified Agent Framework</h4>
<p class="content">
Submit a modified EmbodiedBench agent framework. Teams may customize files under <code>embodiedbench/planner/</code> (for example prompt construction, reasoning, memory, replanning, and action parsing in planner modules such as <code>vlm_planner.py</code> and <code>nav_planner.py</code>).
</p>
<p class="content">
<strong>Constraints:</strong> Teams may only modify files under <code>embodiedbench/planner/</code> and add new supporting modules. Evaluator code, environment code, and metric computation must remain unchanged.
</p>
<p class="content" style="margin-top: 1rem;">
Submit a code zip together with a <code>README</code> that explains how to run your modified framework for Stage 2 evaluation on the held-out EB-ALFRED and EB-Navigation tasks.
</p>
</div>
<div class="challenge-card">
<h4 class="title is-5">Option A + B — Fine-tuned Model with Custom Agent Framework</h4>
<p class="content">
Teams may combine both options by submitting a fine-tuned model served via <a href="https://docs.vllm.ai" target="_blank">vLLM</a> together with a custom planner under <code>embodiedbench/planner/</code> that calls it. This supports end-to-end optimization of both model weights and agent strategy.
</p>
<p class="content">
For the planner portion, the same constraints as Option B apply: teams may only modify files under <code>embodiedbench/planner/</code> and add new supporting modules, while evaluator code, environment code, and metric computation must remain unchanged.
</p>
</div>
<div class="challenge-card">
<h4 class="title is-5">Submission Package</h4>
<ul class="content">
<li>Model weights or code repository (Hugging Face model ID, GitHub link, or compressed archive)</li>
<li>For Option B, a code zip of the modified agent framework</li>
<li>For Option A + B, both the model package or model ID and a code zip of the custom planner framework</li>
<li>A <code>README</code> with step-by-step startup instructions and any customization notes needed for reproduction</li>
<li>Complete dependency list (<code>requirements.txt</code> or equivalent)</li>
<li>Any required <code>vllm serve</code> flags or other startup arguments</li>
<li>For Option B and Option A + B, the submitted package must preserve the original evaluator, environment, and metric code unchanged</li>
</ul>
</div>
</div>
</section>
<!-- Rules -->
<section class="section" id="rules">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Rules</h2>
<ul class="rules-list">
<li>
<span class="rule-icon"><i class="fas fa-times-circle" style="color:#c0392b;"></i></span>
<div>
<strong>Commercial APIs not allowed.</strong> Models must not rely on commercial API calls (e.g., GPT-4, Claude, Gemini) for inference during evaluation.
</div>
</li>
<li>
<span class="rule-icon"><i class="fas fa-check-circle"></i></span>
<div>
<strong>Model size limit:</strong> Open-source base models must have fewer than <strong>10B parameters</strong> to emphasize algorithmic design over model scale.
</div>
</li>
<li>
<span class="rule-icon"><i class="fas fa-check-circle"></i></span>
<div>
<strong>External data and pre-trained models:</strong> Allowed with disclosure. Clearly list all external resources in your submission.
</div>
</li>
<li>
<span class="rule-icon"><i class="fas fa-times-circle" style="color:#c0392b;"></i></span>
<div>
<strong>Human-in-the-loop labeling on test:</strong> Disallowed. Do not attempt to obtain test labels or manipulate evaluation.
</div>
</li>
<li>
<span class="rule-icon"><i class="fas fa-clipboard-check"></i></span>
<div>
<strong>Verification:</strong> Top teams will be asked to provide a technical report and reproducibility details. Stage 2 teams must provide a runnable model or framework.
</div>
</li>
<li>
<span class="rule-icon"><i class="fas fa-users"></i></span>
<div>
<strong>Teams:</strong> Team size is limited to 5 members. Each team may only submit under one team name.
</div>
</li>
</ul>
</div>
</section>
<!-- Baselines & Starter Kit -->
<section class="section" id="baselines">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Baselines & Starter Kit</h2>
<p class="content">
Baselines, environment setup, and evaluation scripts are available in the official EmbodiedBench repository:
</p>
<p class="has-text-centered" style="margin-top: 1rem;">
<a href="https://github.com/EmbodiedBench/EmbodiedBench" class="button is-dark" target="_blank">
<span class="icon"><i class="fab fa-github"></i></span>
<span>github.com/EmbodiedBench/EmbodiedBench</span>
</a>
</p>
<p class="content" style="margin-top: 1rem;">
The repository includes baseline model implementations, the <code>copy_json.py</code> submission helper, and detailed instructions for running evaluations on EB-ALFRED and EB-Navigation.
</p>
</div>
</section>
<!-- Contact -->
<section class="section" id="contact">
<div class="container" style="max-width: 900px;">
<h2 class="title is-3" style="text-align: center; margin-bottom: 2rem;">Contact</h2>
<div class="contact-grid">
<div class="contact-card contact-card--channels">
<h3 class="title is-5">Community Channels</h3>
<p>
Join the EmbodiedBench community on Slack, or use GitHub Issues for technical questions and announcements.
</p>
<div class="contact-actions">
<a href="https://join.slack.com/t/embodiedbench-flx2674/shared_invite/zt-3vtr5ifw3-~o1DFd3JkOjNhtBJ~vie6w" class="button is-dark is-rounded" target="_blank" rel="noopener noreferrer">
<span class="icon"><i class="fab fa-slack"></i></span>
<span>Join Slack</span>
</a>
<a href="https://github.com/EmbodiedBench/EmbodiedBench/issues" class="button is-dark is-rounded" target="_blank" rel="noopener noreferrer">
<span class="icon"><i class="fab fa-github"></i></span>
<span>GitHub Issues</span>
</a>
</div>
</div>
<div class="contact-card contact-card--wechat">
<h3 class="title is-5">WeChat Group</h3>
<p>Scan the QR code below to join the EmbodiedBench WeChat group.</p>
<img src="website/img/wechat-group-qr.jpg" alt="EmbodiedBench WeChat group QR code" class="contact-qr">
<p class="contact-note">The QR code image may expire periodically. If it stops working, please use Slack or GitHub to contact the organizers.</p>
</div>
</div>
<p class="has-text-centered" style="margin-top: 2rem;">
<a href="index.html" class="nav-link-challenge">
<span class="icon"><i class="fas fa-arrow-left"></i></span>
<span>Back to EmbodiedBench Home</span>
</a>
</p>
</div>
</section>
<footer class="footer">
<div class="content has-text-centered">
<p style="font-size: 0.95rem;">
This website is adapted from <a href="https://nerfies.github.io/">Nerfies</a>, 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>
</div>
</footer>
</body>
</html>