Optimized Uncertainty Estimation for Vision Transformers: Enhancing Adversarial Robustness and Performance Using Selective Classification

Authors

Erik Pautsch, Loyola University ChicagoFollow
John Li, University of California - San Diego
Silvio Rizzi, Argonne National LaboratoryFollow
George K. Thiruvathukal, Loyola University ChicagoFollow
Maria Pantoja, California Polytechnic State University, San Luis Obispo

Document Type

Conference Proceeding

Publication Date

11-2023

Publication Title

SC-W '23: Proceedings of the SC '23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis

Pages

391–394

Publisher Name

ACM

Publisher Location

Denver, CO

Abstract

Deep Learning models often exhibit undue confidence when encountering out-of-distribution (OOD) inputs, misclassifying with high confidence. The ideal outcome, in these cases, would be an "I do not know" verdict. We enhance the trustworthiness of our models through selective classification, allowing the model to abstain from making predictions when facing uncertainty. Rather than a singular prediction, the model offers a prediction distribution, enabling users to gauge the model’s trustworthiness and determine the need for human intervention. We assess uncertainty in two baseline models: a Convolutional Neural Network (CNN) and a Vision Transformer (ViT). By leveraging these uncertainty values, we minimize errors by refraining from predictions during high uncertainty. Additionally, we evaluate these models across various distributed architectures, including new AI architectures, Cerebras CS-2, and SambaNova SN30.

Comments

Author Posting © The Authors, 2023. This article is posted here by permission of the ACM for personal use, not for redistribution. The definitive version was published in SC-W '23: Proceedings of the SC '23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis, Pages 391-394, November, 2023. https://doi.org/10.1145/3624062.3624106

Recommended Citation

Pautsch, Erik; Li, John; Rizzi, Silvio; Thiruvathukal, George K.; Pantoja, Maria (2023). Optimized Uncertainty Estimation for Vision Transformers: Enhancing Adversarial Robustness and Performance Using Selective Classification. In proceedings of First Workshop on Enabling Predictive Science with Optimization and Uncertainty Quantification in HPC (EPSOUQ-HPC 2023), https://doi.org/10.6084/m9.figshare.24397216

Creative Commons License

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.

Copyright Statement

© The Authors, 2023.