Enhanced Regularizers for Attributional Robustness

Sarkar, A. and Sarkar, A. and Balasubramanian, Vineeth N (2021) Enhanced Regularizers for Attributional Robustness. In: 35th AAAI Conference on Artificial Intelligence, AAAI 2021, 2 February 2021 through 9 February 2021, Virtual, Online.

Full text not available from this repository. (Request a copy)

Abstract

Deep neural networks are the default choice of learning models for computer vision tasks. Extensive work has been carried out in recent years on explaining deep models for vision tasks such as classification. However, recent work has shown that it is possible for these models to produce substantially different attribution maps even when two very similar images are given to the network, raising serious questions about trustworthiness. To address this issue, we propose a robust attribution training strategy to improve attributional robustness of deep neural networks. Our method carefully analyzes the requirements for attributional robustness and introduces two new regularizers that preserve a model's attribution map during attacks. Our method surpasses state-of-the-art attributional robustness methods by a margin of approximately 3% to 9% in terms of attribution robustness measures on several datasets including MNIST, FMNIST, Flower and GTSRB. © 2021, Association for the Advancement of Artificial Intelligence

[error in script]
IITH Creators:
IITH CreatorsORCiD
Balasubramanian, Vineeth Nhttps://orcid.org/0000-0003-2656-0375
Item Type: Conference or Workshop Item (Paper)
Additional Information: This work has been partly supported by the funding received from MHRD, Govt of India, and Honeywell through the UAY program (UAY/IITH005). We also acknowledge IIT-Hyderabad and JICA for provision of GPU servers for the work. We thank the anonymous reviewers for their valuable feedback that improved the presentation of this work.
Uncontrolled Keywords: Learning models; Regularizer; Robustness measures; Similar image; State of the art; Training strategy
Subjects: Computer science
Divisions: Department of Computer Science & Engineering
Depositing User: . LibTrainee 2021
Date Deposited: 19 Aug 2022 09:06
Last Modified: 19 Aug 2022 09:06
URI: http://raiith.iith.ac.in/id/eprint/10224
Publisher URL:
Related URLs:

Actions (login required)

View Item View Item
Statistics for RAIITH ePrint 10224 Statistics for this ePrint Item