mlsanityData to 14 Oct 2026

Generalizable Image Repair for Robust Visual Control

arXiv 2503.05911 ↗First posted 7 Mar 2025cs.RO cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Vision-based control relies on accurate perception to achieve robustness. However, image distribution changes caused by sensor noise, adverse weather, and dynamic lighting can degrade perception, leading to suboptimal control decisions. Existing approaches, including domain adaptation and adversarial training, improve robustness but struggle to generalize to unseen corruptions while introducing computational overhead. To address this challenge, we propose a real-time image repair module that restores corrupted images before they are used by the controller. Our method leverages generative adversarial models, specifically CycleGAN and pix2pix, for image repair. CycleGAN enables unpaired image-to-image translation to adapt to novel corruptions, while pix2pix exploits paired image data when available to improve the quality. To ensure alignment with control performance, we introduce a control-focused loss function that prioritizes perceptual consistency in repaired images. We evaluated our method in a simulated autonomous racing environment with various visual corruptions. The results show that our approach significantly improves performance compared to baselines, mitigating distribution shift and enhancing controller reliability.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 38 of its 49 references that are in the map. First posted on arXiv 7 Mar 2025.

Builds on

38 of 49 references.

References · 38

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  4. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  5. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  6. 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,527
  7. 2015Domain-Adversarial Training of Neural Networkscited by 3,044
  8. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,626
  9. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,917
  10. 2021SwinIR: Image Restoration Using Swin Transformercited by 1,887
  11. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,517
  12. 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,485
26 more
  1. 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,432
  2. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,340
  3. 2018ESRGAN: Enhanced Super-Resolution Generative Adversarial Networkscited by 1,254
  4. 2014How transferable are features in deep neural networks?cited by 1,225
  5. 2017Deep Image Priorcited by 1,098
  6. 2016Pixel Recurrent Neural Networkscited by 816
  7. 2020Fast is better than free: Revisiting adversarial trainingcited by 727
  8. 2017End-to-end Driving via Conditional Imitation Learningcited by 485
  9. 2020Plug-and-Play Image Restoration with Deep Denoiser Priorcited by 470
  10. 2017Robust Adversarial Reinforcement Learningcited by 451
  11. 2015Return of Frustratingly Easy Domain Adaptationcited by 391
  12. 2021MEMO: Test Time Robustness via Adaptation and Augmentationcited by 377
  13. 2018Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Modelscited by 362
  14. 2018Feature Denoising for Improving Adversarial Robustnesscited by 314
  15. 2020A Brief Review of Domain Adaptationcited by 288
  16. 2017PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examplescited by 244
  17. 2017Deep Bilateral Learning for Real-Time Image Enhancementcited by 200
  18. 2021A Fine-Grained Analysis on Distribution Shiftcited by 171
  19. 2021Towards Faster and Stabilized GAN Training for High-fidelity Few-shot Image Synthesiscited by 140
  20. 2016EPOpt: Learning Robust Neural Network Policies Using Model Ensemblescited by 138
  21. 2017Measuring the tendency of CNNs to Learn Surface Statistical Regularitiescited by 87
  22. 2022Back to the Source: Diffusion-Driven Test-Time Adaptationcited by 71
  23. 2022Domain Adaptive Object Detection for Autonomous Driving under Foggy Weathercited by 41
  24. 2020Bayesian Robust Optimization for Imitation Learningcited by 28
  25. 2021Vision-Based Autonomous Car Racing Using Deep Imitative Reinforcement Learningcited by 21
  26. 2017RAIL: Risk-Averse Imitation Learningcited by 9

Built on by

Not cited yet.