FACL-Attack.github.io/index.html at main · FACL-Attack/FACL-Attack.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
<!DOCTYPE html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="description"
        content="Frequency-Aware Contrastive Learning for Transferable Adversarial Attacks">
  <meta name="keywords" content="FACL-Attack, Transferable Adversarial Attack, Transferable Attack, Transfer Attack, Adversarial Attack">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>FACL-Attack</title>

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

  <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">

  <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>
</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-2 publication-title">Frequency-Aware Contrastive Learning for <br> Transferable Adversarial Attacks</h1>
          <p class="is-size-3"; style="color:#808080; margin-top:-25px"> AAAI 2024 </p>
          <div class="is-size-5 publication-authors">
            <span class="author-block">
              <a href="https://hmyang1.github.io">Hunmin Yang</a><sup>1,2,*</sup>,
            </span>
            <span class="author-block">
              <a href="https://sites.google.com/view/jongohjeong?pli=1">Jongoh Jeong</a><sup>1,*</sup>,
            </span>
            <span class="author-block">
              <a href="https://sites.google.com/site/kjyoon">Kuk-Jin Yoon</a><sup>1</sup>
            </span>
          </div>
          <div class="is-size-5 publication-authors">
            <span class="author-block">
              <sup>1</sup>KAIST,
            </span>
            <span class="author-block">
              <sup>2</sup>ADD
            </span>
          </div>
          <div class="column has-text-centered">
            <div class="publication-links">
              <!-- PDF Link. -->
              <span class="link-block">
                <a href="https://arxiv.org/pdf/2407.20653"
                   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://arxiv.org/abs/2407.20653"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="ai ai-arxiv"></i>
                  </span>
                  <span>arXiv</span>
                </a>
              </span>
              <span class="link-block">
                <a href="#BibTeX"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="ai ai-obp"></i>
                  </span>
                  <span>BibTex</span>
                </a>
              </span>
              <span class="link-block">
                <a href="mailto:facl.attack@gmail.com"
                   class="external-link button is-normal is-rounded is-dark">
                  <span class="icon">
                      <i class="fas fa-envelope"></i>
                  </span>
                  <span>Contact</span>
                </a>
              </span>
            </div>
          </div>
        </div>
      </div>
    </div>
  </div>
</section>


<!-- Teaser -->
<section class="hero teaser" style="margin-top:-40px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static/images/Motivation.png"  class="center"/>
    </div>
    <h2 class="subtitle has-text-centered" style="margin-top:-12px">
    To boost the transferability of adversarial examples, we exploit band-specific characteristics of natural images in the frequency domain.
    Our approach randomizes <i>domain-variant</i> low- and high-band data (FADR module), and perturbs <i>domain-invariant</i> mid-band features (FACL module).
      <br><br>
      <div class="gray-box-custom" style="margin-top:-12px">
        <b>FACL-Attack</b> enhances transferable adversarial attacks via frequency domain manipulations.
      </div>
    </h2>
  </div>
</section>


<!-- Abstract -->
<br>
<hr>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <h2 class="title is-3" style="margin-top:-5px">Abstract</h2>
        <div class="content has-text-justified">
          <p>
            Deep neural networks are known to be vulnerable to security risks due to the inherent transferable nature of adversarial examples.
            Despite the success of recent generative model-based attacks demonstrating strong transferability, it still remains a challenge to design an efficient attack strategy in a real-world strict black-box setting, where both the target domain and model architectures are unknown.
            In this paper, we seek to explore a feature contrastive approach in the frequency domain to generate adversarial examples that are robust in both cross-domain and cross-model settings.
            With that goal in mind, we propose two modules that are only employed during the training phase: a <b>F</b>requency-<b>A</b>ware <b>D</b>omain <b>R</b>andomization (FADR) module to randomize domain-variant low- and high-range frequency components and a <b>F</b>requency-<b>A</b>ugmented <b>C</b>ontrastive <b>L</b>earning (FACL) module to effectively separate domain-invariant mid-frequency features of clean and perturbed image.
            We demonstrate strong transferability of our generated adversarial perturbations through extensive cross-domain and cross-model experiments, while keeping the inference time complexity.
          </p>
        </div>
      </div>
    </div>
</section>


<!-- Method -->
<br>
<hr>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <h2 class="title is-3" style="margin-top:-5px">Method</h2>
      </div>
    </div>
</section>


<br>
<section class="hero teaser" style="margin-top:-5px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static/images/Overview.png" class="center-img"/>
    </div>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div class="content has-text-justified" style="margin-top:-22px">
          <p>
            <b>Overview of FACL-Attack.</b>&nbsp;
            From the clean input image, our FADR module outputs the augmented image after spectral transformation, which is targeted to randomize only the domain-variant low/high FCs.
            The perturbation generator then produces the bounded adversarial image with perturbation projector from the randomized image.
            The resulting clean and adversarial image pairs are decomposed into mid-band (<i>domain-agnostic</i>) and low/high-band (<i>domain-specific</i>) FCs, whose features extracted from the k-th middle layer of the surrogate model are contrasted in our FACL module to boost the adversarial transferability.
            The adversarial image is colorized for visualization.
          </p>
        </div>
      </div>
    </div>
</section>


<br>
<section class="hero teaser" style="margin-top:-5px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static/images/FADR_Image.png" class="center-img"/>
    </div>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div class="content has-text-justified" style="margin-top:-22px">
          <p>
            <b>Visualization of spectral transformation by FADR.</b>&nbsp;
            From the clean input image (column 1), our FADR decomposes the image into mid-band (column 2) and low/high-band (column 3) FCs.
            The FADR only randomizes the <i>domain-variant</i> low/high-band FCs, yielding the augmented output in column 4.
            Here we demonstrate transformations with large hyper-parameters for visualization.
          </p>
        </div>
      </div>
    </div>
</section>


<br>
<section class="hero teaser" style="margin-top:-5px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static/images/FACL_Difference_Map.png" class="center-img"/>
    </div>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div class="content has-text-justified" style="margin-top:-22px">
          <p>
            <b>Difference map of perturbed features by FACL.</b>&nbsp;
            Clean image, unbounded adversarial images from baseline and FACL, and the final difference map (Diff(baseline, baseline+FACL)), from left to right.
            Our generated perturbations are more focused on domain-agnostic semantic region such as shape, facilitating more transferable attack.
          </p>
        </div>
      </div>
    </div>
</section>


<!-- Experimental Results -->
<br>
<hr>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <h2 class="title is-3" style="margin-top:-5px">Experimental Results</h2>
      </div>
    </div>
</section>


<br>
<section class="hero teaser" style="margin-top:-5px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static/images/Experimental_Results_Cross_Domain.png" class="center-img"/>
    </div>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div class="content has-text-justified" style="margin-top:-22px">
          <p>
            <b>Cross-domain evaluation results.</b>&nbsp;
            The perturbation generator is trained on ImageNet-1K with VGG-16 as the surrogate model and evaluated on black-box domains with models.
            We compare the top-1 classification accuracy after attacks.
          </p>
        </div>
      </div>
    </div>
</section>


<br><br>
<section class="hero teaser" style="margin-top:-5px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static/images/Experimental_Results_Cross_Model.png" class="center-img"/>
    </div>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div class="content has-text-justified" style="margin-top:-22px">
          <p>
            <b>Cross-model evaluation results.</b>&nbsp;
            The perturbation generator is trained on ImageNet-1K with VGG-16 as the surrogate model and evaluated on black-box models.
            We compare the top-1 classification accuracy after attacks.
          </p>
        </div>
      </div>
    </div>
</section>


<br><br>
<section class="hero teaser" style="margin-top:-5px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static/images/Experimental_Results_Cross_Model_SOTA.png" class="center-img"/>
    </div>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div class="content has-text-justified" style="margin-top:-22px">
          <p>
            <b>Evaluation on the state-of-the-art models.</b>&nbsp;
            The perturbation generator is trained on ImageNet-1K with VGG-16 as the surrogate model and evaluated on black-box models.
            We compare the top-1 classification accuracy after attacks.
          </p>
        </div>
      </div>
    </div>
</section>


<br>
<section class="hero teaser" style="margin-top:20px">
  <div class="container is-max-desktop">
    <div class="hero-body">
      <img src="static\images\Qualtitative_Results.png" class="center-img"/>
    </div>
  </div>
</section>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <div class="content has-text-justified" style="margin-top:-22px">
          <p>
            <b>Qualitative results on various domains.</b>&nbsp;
            FACL-Attack successfully fools the classifier, causing it to predict the clean image labels (in black) as the mispredicted class labels shown at the bottom (in red).
            From top to bottom: clean images, unbounded adversarial images, and bounded adversarial images which are <i>actual inputs</i> to the classifier.
          </p>
        </div>
      </div>
    </div>
</section>


<!-- Contact. -->
<br>
<hr>
<section class="section">
  <div class="container is-max-desktop">
    <div class="columns is-centered has-text-centered">
      <div class="column is-four-fifths">
        <h2 class="title is-3" style="margin-top:-5px">Contact</h2>
        <div class="is-centered has-text-centered is-size-5">
          <p>
            FACL-Attack (<a href="mailto:facl.attack@gmail.com">facl.attack@gmail.com</a>)
          </p>
        </div>
      </div>
    </div>
</section>


<!-- BibTex. -->
<hr>
<section class="section" id="BibTeX">
  <div class="container is-max-desktop content">
    <h2 class="title">BibTex</h2>
    <pre><code>@InProceedings{yang2024FACLAttack,
      title={Frequency-Aware Contrastive Learning for Transferable Adversarial Attacks},
      author={Hunmin Yang and Jongoh Jeong and Kuk-Jin Yoon},
      booktitle={AAAI},
      year={2024}
}</code></pre>
  </div>
</section>
<br><br>

<footer class="footer">
  <div class="container">
    <div class="content has-text-centered">
      <a class="icon-link"
         href="https://arxiv.org/pdf/2407.20653">
        <i class="fas fa-file-pdf"></i>
      </a>
      <a class="icon-link" href="https://arxiv.org/abs/2407.20653" class="external-link" disabled>
        <i class="fas fa-user"></i>
      </a>
      <a class="icon-link" href="mailto:facl.attack@gmail.com" class="external-link" disabled>
        <i class="fas fa-envelope"></i>
      </a>
    </div>
    <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>
            This website template was adapted from <a href="https://nerfies.github.io/">Nerfies website</a>.
          </p>
        </div>
      </div>
    </div>
  </div>
</footer>

</body>
</html>