Please use this identifier to cite or link to this item:
Title: CNN fixations : an unraveling approach to visualize the discriminative image regions
Authors: Mopuri, Konda Reddy
Garg, Utsav
Babu, R. Venkatesh
Keywords: Engineering::Computer science and engineering
Issue Date: 2018
Source: Mopuri, K. R., Garg, U., & Babu, R. V. (2019). CNN fixations : an unraveling approach to visualize the discriminative image regions. IEEE Transactions on Image Processing, 28(5), 2116-2125. doi:10.1109/TIP.2018.2881920
Journal: IEEE Transactions on Image Processing
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.
ISSN: 1057-7149
DOI: 10.1109/TIP.2018.2881920
Rights: © 2018 IEEE. All rights reserved.
Fulltext Permission: none
Fulltext Availability: No Fulltext
Appears in Collections:SCSE Journal Articles

Citations 10

Updated on Mar 25, 2023

Web of ScienceTM
Citations 10

Updated on Mar 20, 2023

Page view(s)

Updated on Mar 26, 2023

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.