mlsanityData to 7 Oct 2026

ROPUST: Improving Robustness through Fine-tuning with Photonic Processors and Synthetic Gradients

arXiv 2108.04217 ↗First posted 6 Jul 2021cs.CV cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

Robustness to adversarial attacks is typically obtained through expensive adversarial training with Projected Gradient Descent. Here we introduce ROPUST, a remarkably simple and efficient method to leverage robust pre-trained models and further increase their robustness, at no cost in natural accuracy. Our technique relies on the use of an Optical Processing Unit (OPU), a photonic co-processor, and a fine-tuning step performed with Direct Feedback Alignment, a synthetic gradient training scheme. We test our method on nine different models against four attacks in RobustBench, consistently improving over state-of-the-art performance. We perform an ablation study on the single components of our defense, showing that robustness arises from parameter obfuscation and the alternative training method. We also introduce phase retrieval attacks, specifically designed to increase the threat level of attackers against our own defense. We show that even with state-of-the-art phase retrieval techniques, ROPUST remains an effective defense.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 46 of its 50 references that are in the map. First posted on arXiv 6 Jul 2021.

Builds on

46 of 50 references.

References · 46

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  3. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  4. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
  5. 2016Identity Mappings in Deep Residual Networkscited by 1,960
  6. 2016Adversarial examples in the physical worldcited by 1,762
  7. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,442
  8. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,324
  9. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,259
  10. 2016Practical Black-Box Attacks against Machine Learningcited by 893
  11. 2020Fast is better than free: Revisiting adversarial trainingcited by 723
  12. 2019Square Attack: a query-efficient black-box adversarial attack via random searchcited by 682
34 more
  1. 2020RobustBench: a standardized adversarial robustness benchmarkcited by 659
  2. 2017ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Modelscited by 656
  3. 2017Provable defenses against adversarial examples via the convex outer adversarial polytopecited by 564
  4. 2018Black-box Adversarial Attacks with Limited Queries and Informationcited by 499
  5. 2018Certified Robustness to Adversarial Examples with Differential Privacycited by 462
  6. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 439
  7. 2019Unlabeled Data Improves Adversarial Robustnesscited by 415
  8. 2019Using Pre-Training Can Improve Model Robustness and Uncertaintycited by 373
  9. 2019Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attackcited by 333
  10. 2020Do Adversarially Robust ImageNet Models Transfer Better?cited by 315
  11. 2020Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examplescited by 275
  12. 2016Direct Feedback Alignment Provides Learning in Deep Neural Networkscited by 240
  13. 2020Geometry-aware Instance-reweighted Adversarial Trainingcited by 198
  14. 2019Adversarial Training and Robustness for Multiple Perturbationscited by 182
  15. 2018MMA Training: Direct Input Space Margin Maximization through Adversarial Trainingcited by 171
  16. 2020The Hardware Lotterycited by 160
  17. 2018Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priorscited by 145
  18. 2018Scaling provable adversarial defensescited by 141
  19. 2018Adversarial examples from computational constraintscited by 79
  20. 2019Adversarially robust transfer learningcited by 66
  21. 2020Direct Feedback Alignment Scales to Modern Deep Learning Tasks and Architecturescited by 56
  22. 2019Parsimonious Black-Box Adversarial Attacks via Efficient Combinatorial Optimizationcited by 49
  23. 2014Random feedback weights support learning in deep neural networkscited by 47
  24. 2019Theoretical evidence for adversarial robustness through randomizationcited by 33
  25. 2020A Simple Fine-tuning Is All You Need: Towards Robust Deep Learning Via Adversarial Fine-tuningcited by 30
  26. 2019Yet another but more efficient black-box adversarial attack: tiling and evolution strategiescited by 17
  27. 2020Advocating for Multiple Defense Strategies against Adversarial Examplescited by 11
  28. 2021Adversarial Robustness by Design through Analog Computing and Synthetic Gradientscited by 7
  29. 2019Kernel computations from large-scale random features obtained by Optical Processing Unitscited by 6
  30. 2019Adversarial Robustness via Label-Smoothingcited by 5
  31. 2019On the Adversarial Robustness of Neural Networks without Weight Transportcited by 5
  32. 2019Don't take it lightly: Phasing optical random projections with unknown operatorscited by 4
  33. 2019Fast Optical System Identification by Numerical Interferometrycited by 4
  34. 2020On the Intrinsic Robustness of NVM Crossbars Against Adversarial Attackscited by 2

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 1.

2 papers.

  1. 2021Benchmarking the Accuracy and Robustness of Feedback Alignment Algorithmscited by 9
  2. 2022Scaling Laws Beyond Backpropagationcited by 4