2021 IEEE International Conference on Image Processing

19-22 September 2021 • Anchorage, Alaska, USA

Imaging Without Borders

2021 IEEE International Conference on Image Processing

19-22 September 2021 • Anchorage, Alaska, USA

Login Paper Search My Schedule Paper Index Help

My ICIP 2021 Schedule

Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.
  1. Create a login based on your email (takes less than one minute)
  2. Perform 'Paper Search'
  3. Select papers that you desire to save in your personalized schedule
  4. Click on 'My Schedule' to see the current list of selected papers
  5. Click on 'Printable Version' to create a separate window suitable for printing (the header and menu will appear, but will not actually print)

Paper Detail

Paper IDMLR-APPL-IVSMR-2.8
Paper Title RETHINKING TRAINING SCHEDULES FOR VERIFIABLY ROBUST NETWORKS
Authors Hyojun Go, Junyoung Byun, Changick Kim, Korea Advanced Institute of Science and Technology, Republic of Korea
SessionMLR-APPL-IVSMR-2: Machine learning for image and video sensing, modeling and representation 2
LocationArea D
Session Time:Tuesday, 21 September, 15:30 - 17:00
Presentation Time:Tuesday, 21 September, 15:30 - 17:00
Presentation Poster
Topic Applications of Machine Learning: Machine learning for image & video sensing, modeling, and representation
IEEE Xplore Open Preview  Click here to view in IEEE Xplore
Abstract New and stronger adversarial attacks can threaten existing defenses. This possibility highlights the importance of certified defense methods that train deep neural networks with verifiably robust guarantees. A range of certified defense methods has been proposed to train neural networks with verifiably robustness guarantees, among which Interval Bound Propagation (IBP) and CROWN-IBP have been demonstrated to be the most effective. However, we observe that CROWN-IBP and IBP are suffering from Low Epsilon Overfitting (LEO), a problem arising from their training schedule that increases the input perturbation bound. We show that LEO can yield poor results even for a simple linear classifier. We also investigate the evidence of LEO from experiments under conditions of worsening LEO. Based on these observations, we propose a new training strategy, BatchMix, which mixes various input perturbation bounds in a mini-batch to alleviate the LEO problem. Experimental results on MNIST and CIFAR-10 datasets show that BatchMix can make the performance of IBP and CROWN-IBP better by mitigating LEO.