dc.citation.conferencePlace |
ZZ |
- |
dc.citation.conferencePlace |
Online |
- |
dc.citation.title |
Neural Information Processing Systems |
- |
dc.contributor.author |
Lee, Sungyoon |
- |
dc.contributor.author |
Lee, Jaewook |
- |
dc.contributor.author |
Park, Saerom |
- |
dc.date.accessioned |
2024-01-31T22:09:33Z |
- |
dc.date.available |
2024-01-31T22:09:33Z |
- |
dc.date.created |
2023-05-30 |
- |
dc.date.issued |
2020-12-06 |
- |
dc.description.abstract |
Verifiable training is a promising research direction for training a robust network. However, most verifiable training methods are slow or lack scalability. In this study, we propose a fast and scalable certifiable training algorithm based on Lipschitz analysis and interval arithmetic. Our certifiable training algorithm provides a tight propagated outer bound by introducing the box constraint propagation (BCP), and it efficiently computes the worst logit over the outer bound. In the experiments, we show that BCP achieves a tighter outer bound than the global Lipschitz-based outer bound. Moreover, our certifiable training algorithm is over 12 times faster than the state-of-the-art dual relaxation-based method; however, it achieves comparable or better verification performance, improving natural accuracy. Our fast certifiable training algorithm with the tight outer bound can scale to Tiny ImageNet with verification accuracy of 20.1% |
- |
dc.identifier.bibliographicCitation |
Neural Information Processing Systems |
- |
dc.identifier.scopusid |
2-s2.0-85104220603 |
- |
dc.identifier.uri |
https://scholarworks.unist.ac.kr/handle/201301/77738 |
- |
dc.language |
영어 |
- |
dc.publisher |
Neural information processing systems foundation |
- |
dc.title |
Lipschitz-certifiable training with a tight outer bound |
- |
dc.type |
Conference Paper |
- |
dc.date.conferenceDate |
2020-12-06 |
- |