Current browse context:
cs.CV
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Robustness and Visual Explanation for Black Box Image, Video, and ECG Signal Classification with Reinforcement Learning
(Submitted on 27 Mar 2024 (v1), last revised 22 Apr 2024 (this version, v2))
Abstract: We present a generic Reinforcement Learning (RL) framework optimized for crafting adversarial attacks on different model types spanning from ECG signal analysis (1D), image classification (2D), and video classification (3D). The framework focuses on identifying sensitive regions and inducing misclassifications with minimal distortions and various distortion types. The novel RL method outperforms state-of-the-art methods for all three applications, proving its efficiency. Our RL approach produces superior localization masks, enhancing interpretability for image classification and ECG analysis models. For applications such as ECG analysis, our platform highlights critical ECG segments for clinicians while ensuring resilience against prevalent distortions. This comprehensive tool aims to bolster both resilience with adversarial training and transparency across varied applications and data types.
Submission history
From: Soumyendu Sarkar [view email][v1] Wed, 27 Mar 2024 20:07:39 GMT (488kb)
[v2] Mon, 22 Apr 2024 14:49:36 GMT (480kb)
Link back to: arXiv, form interface, contact.