Efficient Adversarial Contrastive Learning via Robustness-Aware Coreset Selection
Dec 10, 2023
Speakers
Jingfeng Zhang
Speaker · 0 followers
Masashi Sugiyama
Speaker · 2 followers
Mohan Kankanhalli
Speaker · 0 followers
About
Adversarial contrastive learning (ACL) does not require expensive data annotations but outputs a robust representation that withstands adversarial attacks and also generalizes to a wide range of downstream tasks. However, ACL needs tremendous running time to generate the adversarial variants of all training data, which limits its scalability to large datasets. To speed up ACL, this paper proposes a robustness-aware coreset selection (RCS) method. RCS does not require label information and searches for an informative subset that minimizes a representational divergence, which is the distance of the representation between natural data and their virtual adversarial variants. The vanilla solution of RCS via traversing all possible subsets is computationally prohibitive. Therefore, we theoretically transform RCS into a surrogate problem of submodular maximization, of which the greedy search is an efficient solution with an optimality guarantee for the original problem. Empirically, our comprehensive results corroborate that RCS can speed up ACL by a large margin without significantly hurting the robustness transferability. Notably, to the best of our knowledge, we are the first to conduct ACL efficiently on the large-scale ImageNet-1K dataset to obtain an effective robust representation via RCS. The source code is available at the anonymous Github.Adversarial contrastive learning (ACL) does not require expensive data annotations but outputs a robust representation that withstands adversarial attacks and also generalizes to a wide range of downstream tasks. However, ACL needs tremendous running time to generate the adversarial variants of all training data, which limits its scalability to large datasets. To speed up ACL, this paper proposes a robustness-aware coreset selection (RCS) method. RCS does not require label information and search…
Organizer
NeurIPS 2023
Account · 623 followers
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
On permutation symmetries in Bayesian neural network posteriors
Watch later
Simone Rossi, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Learning Curves for Deep Structured Gaussian Feature Models
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Cheaply Evaluating Inference Efficiency Metrics for Autoregressive Transformer Models
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Depthwise Hyperparameter Transfer: Dynamics and Scaling Limit
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
On kernel-based statistical learning in the mean field limit
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Learning To Dive In Branch And Bound
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%