CNN Fixations: An Unraveling Approach to Visualize the Discriminative Image Regions

Mopuri, Konda Reddy and Garg, Utsav and Babu, Venkatesh R (2019) CNN Fixations: An Unraveling Approach to Visualize the Discriminative Image Regions. In: IEEE TRANSACTIONS ON IMAGE PROCESSING, 28 (4). pp. 2116-2125.

PDF Iee_Tra_Ima_Pro_28-4_2116_2019.pdf - Published Version Restricted to Registered users only Download (2MB) | Request a copy

Abstract

Deep convolutional neural networks (CNNs) have revolutionized the computer vision research and have seen unprecedented adoption for multiple tasks, such as classification, detection, and caption generation. However, they offer little transparency into their inner workings and are often treated as black boxes that deliver excellent performance. In this paper, we aim at alleviating this opaqueness of CNNs by providing visual explanations for the network's predictions. Our approach can analyze a variety of CNN-based models trained for computer vision applications, such as object recognition and caption generation. Unlike the existing methods, we achieve this via unraveling the forward pass operation. The proposed method exploits feature dependencies across the layer hierarchy and uncovers the discriminative image locations that guide the network's predictions. We name these locations CNN fixations, loosely analogous to human eye fixations. Our approach is a generic method that requires no architectural changes, additional training, or gradient computation, and computes the important image locations (CNN fixations). We demonstrate through a variety of applications that our approach is able to localize the discriminative image locations across different network architectures, diverse vision tasks, and data modalities.

Item Type:	Journal Article
Publication:	IEEE TRANSACTIONS ON IMAGE PROCESSING
Publisher:	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
Additional Information:	Copyright of this article belongs to IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
Keywords:	Explainable AI; CNN visualization; visual explanations; label localization; weakly supervised localization
Department/Centre:	Division of Interdisciplinary Sciences > Computational and Data Sciences
Date Deposited:	12 Feb 2019 04:43
Last Modified:	12 Feb 2019 04:43
URI:	http://eprints.iisc.ac.in/id/eprint/61684

Actions (login required)

View Item