-
Notifications
You must be signed in to change notification settings - Fork 1
/
index.html
392 lines (386 loc) · 24.7 KB
/
index.html
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
<!doctype html>
<html lang="en">
<head>
<!-- Required meta tags -->
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<!-- Bootstrap CSS -->
<link href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css" rel="stylesheet"
integrity="sha384-giJF6kkoqNQ00vy+HMDP7azOuL0xtbfIcaT9wjKHr8RbDVddVHyTfAAsrekwKmP1" crossorigin="anonymous">
<script async src="https://www.googletagmanager.com/gtag/js?id=G-0SENLZZTTL"></script>
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-H4V9YKMB9P"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag() { dataLayer.push(arguments); }
gtag('js', new Date());
gtag('config', 'G-H4V9YKMB9P');
</script>
<link rel="icon" type="image/png" href="images/y-100.png" />
<link href="academicons.min.css" rel="stylesheet">
<link href="all.min.css" rel="stylesheet">
<link href="style.css" rel="stylesheet">
<title>Yossi Gandelsman</title>
</head>
<body>
<div class="intro px-3 py-3 pt-md-5 pb-md-4 mx-auto">
<h1 class="display-6 text-center">Yossi Gandelsman</h1>
<div class="d-flex flex-wrap bd-highlight mb-3 my-3 justify-content-center">
<div class="p-2 order-1 bd-highlight"><img class="img-thumbnail" src="images/me.png" alt="" width="240px">
</div><div class="intro-text order-2 p-2 my-1 bd-highlight text-left">
<p>I am a Ph.D. student at <a href="https://bair.berkeley.edu/">Berkeley AI Research</a> (UC Berkeley), advised by <a
href="https://people.eecs.berkeley.edu/~efros/">Prof. Alexei Efros</a>, and a visiting researcher at <a href="https://ai.meta.com/research/">FAIR</a> (Meta). I work on computer
vision and deep learning problems. I am funded by the <a href='https://research.google/outreach/phd-fellowship/recipients/'>Google Ph.D. Fellowship</a>.</p>
<p>Previously, I was a member of the Perception Team at <a target="_blank"
href="https://ai.google/research/">Google Research</a> (now Google-DeepMind).</p>
<p>I completed my M.Sc. at the faculty of <a target="_blank" href="http://www.weizmann.ac.il/math/">
Mathematics and Computer Science of the Weizmann Institute of Science</a>, advised by
<a target="_blank" href="http://www.weizmann.ac.il/math/irani/">Prof. Michal Irani</a>, and a B.Sc. at the <a href="https://www.openu.ac.il/en/pages/default.aspx">Open University of Israel</a>.</p>
<br>
<p class="text-center"><a target="_blank" href="mailto:[email protected]"><i
class="far fa-envelope"></i></a>
<!--<a target="_blank" target="_blank" href="data/CV.pdf"><i class="ai ai-cv"></i></a> --><a
target="_blank" href="https://scholar.google.co.il/citations?user=71L4yYMAAAAJ&hl=en"><i
class="ai ai-google-scholar"></i></a> <a target="_blank"
href="https://www.linkedin.com/in/yossi-gandelsman-b3645a194/"><i
class="fab fa-linkedin"></i></a> <a target="_blank"
href="https://github.com/yossigandelsman/"><i class="fab fa-github"></i></a> <a target="_blank"
href="https://twitter.com/YGandelsman"><i class="fab fa-twitter"></i></a>
</p>
</div>
</div>
</div>
<div class="container">
<a name="publications"></a>
<h2 class="display-7 mb-2 text-center">Selected Publications</h2>
<p class="text-center">Please visit <a target="_blank" href="https://scholar.google.co.il/citations?user=71L4yYMAAAAJ&hl=en">my google scholar</a> for the full list of publications</p>
<br>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://unicorn53547.github.io/video_syn_rep/">
<img src="images/videomae.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Learning Video Representations without Natural Videos</h4>
<ul class="list-unstyled mb-0">
<li>Xueyang Yu, Xinlei Chen and <b>Yossi Gandelsman</b></li>
<li><i>Preprint</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2410.24213">paper</a> | <a
target="_blank" href="https://unicorn53547.github.io/video_syn_rep/">project page</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://anishk23733.github.io/vl-interp/">
<img src="images/vlm_hallucinations.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Interpreting and Editing Vision-Language Representations<br>to Mitigate Hallucinations</h4>
<ul class="list-unstyled mb-0">
<li>Nick Jiang*, Anish Kachinthaya*, Suzie Petryk† and <b>Yossi Gandelsman†</b></li>
<li><i>Preprint</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2410.02762">paper</a> | <a
target="_blank" href="https://anishk23733.github.io/vl-interp/">project page</a> | <a href="https://github.com/nickjiang2378/vl-interp">code</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="clip_neurons/index.html">
<img src="images/clip_neurons.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Interpreting the Second-Order Effects of Neurons in CLIP</h4>
<ul class="list-unstyled mb-0">
<li><b>Yossi Gandelsman</b>, Alexei A. Efros and Jacob Steinhardt</li>
<li><i>Preprint</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2406.04341">paper</a> | <a
target="_blank" href="clip_neurons/index.html">project page</a> | <a href="https://github.com/yossigandelsman/second_order_lens">code</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://snap-research.github.io/weights2weights/">
<img src="images/w2w.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Interpreting the Weight Space of Customized Diffusion Models</h4>
<ul class="list-unstyled mb-0">
<li>Amil Dravid*, <b>Yossi Gandelsman*</b>, Kuan-Chieh Wang, Rameen Abdal, Gordon Wetzstein, Alexei A. Efros and Kfir Aberman</li>
<li><i>NeurIPS 2024</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2406.09413">paper</a> | <a
target="_blank" href="https://snap-research.github.io/weights2weights">project page</a> | <a href="https://github.com/snap-research/weights2weights">code</a> | <a href="https://huggingface.co/snap-research/weights2weights">dataset</a> | <a href="https://huggingface.co/spaces/snap-research/weights2weights">demo</a> | <a href="https://www.youtube.com/watch?v=95raWv_k08c&ab_channel=AmilDravid">video</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://sap3d.github.io/">
<img src="images/sap3d.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">The More You See in 2D, the More You Perceive in 3D</h4>
<ul class="list-unstyled mb-0">
<li>Xinyang Han*, Zelin Gao*, Angjoo Kanazawa, Shubham Goel† and <b>Yossi Gandelsman†</b></li>
<li><i>CVPR 2024 <span class="achievement">[Spotlight]</span></i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2404.03652">paper</a> | <a
target="_blank" href="https://sap3d.github.io/">project page</a> | <a href="https://github.com/sap3d/Sap3D">code</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="clip_decomposition/index.html">
<img src="images/clip_decomposition.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Interpreting CLIP's Image Representation via <br>Text-Based Decomposition</h4>
<ul class="list-unstyled mb-0">
<li><b>Yossi Gandelsman</b>, Alexei A. Efros and Jacob Steinhardt</li>
<li><i>ICLR 2024 <span class="achievement">[Oral]</span></i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2310.05916">paper</a> | <a
target="_blank" href="clip_decomposition/index.html">project page</a> | <a href="https://github.com/yossigandelsman/clip_prs">code</a> | <a href="https://intellabs.github.io/multimodal_cognitive_ai/CLIP-InterpreT/">CLIP-InterpreT (demo)</a> | <a href="https://www.youtube.com/watch?v=l6IVjBvVc_E">video</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="rosetta_neurons/index.html">
<img src="images/rosetta_neurons.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Rosetta Neurons: Mining the Common Units in a Model Zoo</h4>
<ul class="list-unstyled mb-0">
<li>Amil Dravid*, <b>Yossi Gandelsman*</b>, Alexei A. Efros and Assaf Shocher</li>
<li><i>ICCV 2023</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2306.09346">paper</a> | <a
target="_blank" href="rosetta_neurons/index.html">project page</a> | <a href="https://github.com/yossigandelsman/rosetta_neurons">code</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="ttt_mae/index.html">
<img src="images/ttt_mae.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Test-Time Training with Masked Autoencoders</h4>
<ul class="list-unstyled mb-0">
<li><b>Yossi Gandelsman*</b>, Yu Sun*, Xinlei Chen and Alexei A. Efros</li>
<li><i>NeurIPS 2022</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2209.07522">paper</a> | <a
target="_blank" href="ttt_mae/index.html">project page</a> | <a href="https://github.com/yossigandelsman/test_time_training_mae">code</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="visual_prompt/index.html">
<img src="images/visual_prompt.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Visual Prompting via Image Inpainting</h4>
<ul class="list-unstyled mb-0">
<li>Amir Bar*, <b>Yossi Gandelsman*</b>, Trevor Darrell, Amir Globerson and Alexei A. Efros</li>
<li><i>NeurIPS 2022</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2209.00647">paper</a> | <a
target="_blank" href="visual_prompt/index.html">project page</a> | <a href="https://github.com/amirbar/visual_prompting">code</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://mystyle-personalized-prior.github.io/">
<img src="images/my_style.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">MyStyle: A Personalized Generative Prior</h4>
<ul class="list-unstyled mb-0">
<li>Yotam Nitzan, Kfir Aberman, Qiurui He, Orly Liba, Michal Yarom, <b>Yossi Gandelsman</b>, Inbar Mosseri, Yael Pritch and Daniel Cohen-Or</li>
<li><i>SIGGRAPH Asia 2022 (journal track)</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2203.17272">paper</a> | <a
target="_blank" href="https://mystyle-personalized-prior.github.io/">project page</a> | <a href="https://www.youtube.com/watch?v=axWo_9Gt47o&ab_channel=kfiraberman">video</a></a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://dino-vit-features.github.io/">
<img src="images/vit_features.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Deep ViT Features as Dense Visual Descriptors</h4>
<ul class="list-unstyled mb-0">
<li>Shir Amir,
<b>Yossi Gandelsman</b>, Shai Bagon and Tali Dekel
</li>
<li><i>ECCVW 2022 <span class="achievement">["WIMF" Best Spotlight Presentation]</span></i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2112.05814">paper</a> | <a
target="_blank" href="https://dino-vit-features.github.io/">project page</a> | <a
target="_blank" href="https://github.com/ShirAmir/dino-vit-features">code</a></li>
</ul>
</div>
</div>
</div>
</div>
<!-- <div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://deep-saliency-prior.github.io/">
<img src="images/saliency_reduction.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Deep Saliency Prior for Reducing Visual Distraction</h4>
<ul class="list-unstyled mb-0">
<li>Kfir Aberman*,
Junfeng He*, <b>Yossi Gandelsman</b>, Inbar Mosseri, David E. Jacobs, Kai Kohlhoff, Yael Pritch and Michael Rubinstein
</li>
<li><i>CVPR 2022</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2109.01980">paper</a> | <a
target="_blank" href="https://deep-saliency-prior.github.io/">project page</a></li>
</ul>
</div>
</div>
</div>
</div> -->
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://explaining-in-style.github.io/">
<img src="images/explaininginstyle.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Explaining in Style: <br>Training a GAN to explain a
classifier in StyleSpace</h4>
<ul class="list-unstyled mb-0">
<li>Oran Lang*,
<b>Yossi Gandelsman*</b>, Michal
Yarom*, Yoav Wald*, Gal Elidan, Avinatan Hassidim, William T. Freeman, Phillip Isola, Amir Globerson, Michal Irani and Inbar
Mosseri</a>
</li>
<li><i>ICCV 2021</i></li>
<li class="mt-2"><a target="_blank" href="https://arxiv.org/abs/2104.13369">paper</a> | <a
target="_blank" href="https://explaining-in-style.github.io/">project page</a> | <a
target="_blank"
href="https://github.com/google/explaining-in-style">code</a> | <a
target="_blank" href="https://www.youtube.com/watch?v=wLk2eBdXH4M&ab_channel=InbarMosseri">video</a> | <a
target="_blank" href="https://ai.googleblog.com/2022/01/introducing-stylex-new-approach-for.html">blog post</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="https://semantic-pyramid.github.io/">
<img src="images/semanticpyramid.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Semantic Pyramid for Image Generation</h4>
<ul class="list-unstyled mb-0">
<li>Assaf
Shocher*, <b>Yossi Gandelsman*</b>, Inbar
Mosseri, Michal
Yarom, Michal
Irani, William T. Freeman and Tali Dekel</li>
<li><i>CVPR 2020 <span class="achievement">[Oral]</span></i></li>
<li class="mt-2"><a target="_blank"
href="https://semantic-pyramid.github.io/paper.pdf">paper</a> | <a target="_blank"
href="https://semantic-pyramid.github.io/">project page</a> | <a target="_blank"
href="https://www.youtube.com/watch?v=IXtqiJ7Nufo&feature=emb_logo">video</a></li>
</ul>
</div>
</div>
</div>
</div>
<div class="card-deck mb-3">
<div class="card mb-2 shadow-sm">
<div class="card-body p-2 d-flex flex-wrap">
<div class="teaser order-1 p-2 bd-highlight">
<a target="_blank" href="http://www.wisdom.weizmann.ac.il/~vision/DoubleDIP/index.html">
<img src="images/doubledip.png" alt="" width="100%">
</a>
</div>
<div class="p-2 order-2 bd-highlight paper">
<h4 class="my-0 mb-2 font-weight-normal">Double-DIP: Unsupervised Image Decomposition via <br>
Coupled Deep-Image-Priors</h4>
<ul class="list-unstyled mb-0">
<li><b>Yossi Gandelsman</b>, Assaf Shocher and Michal Irani</li>
<li><i>CVPR 2019 <span class="achievement">[Oral]</span></i></li>
<li class="mt-2"><a target="_blank"
href="http://www.wisdom.weizmann.ac.il/~vision/DoubleDIP/resources/DoubleDIP.pdf">paper</a>
| <a target="_blank"
href="http://www.wisdom.weizmann.ac.il/~vision/DoubleDIP/index.html">project
page</a> | <a target="_blank"
href="https://github.com/yossigandelsman/DoubleDIP">code</a> | <a target="_blank"
href="https://youtu.be/IyPQqqGPHws?t=1429">video</a></li>
</ul>
</div>
</div>
</div>
</div>
</div>
</body>
</html>