Understanding convolutional neural networks via discriminant feature analysis
Authors:
Open Access
- 11 December 2018
- journal article
- research article
- Published by Emerald Publishing in APSIPA Transactions on Signal and Information Processing
- Vol. 7 (1)
Abstract
Trained features of a convolution neural network (CNN) at different convolution layers is analyzed using two quantitative metrics in this work. We first show mathematically that the Gaussian confusion measure (GCM) can be used to identify the discriminative ability of an individual feature. Next, we generalize this idea, introduce another measure called the cluster purity measure (CPM), and use it to analyze the discriminative ability of multiple features jointly. The discriminative ability of trained CNN features is validated by experimental results. Research on CNNs utilizing GCM and CPM tools offers important insights into its operational mechanism, including the behavior of trained CNN features and good detection performance of some object classes that were considered difficult in the past. Finally, the trained feature representation is compared between different CNN structures to explain the superiority of deeper networks.Keywords
This publication has 24 references indexed in Scilit:
- Deformable part models are convolutional neural networks2015
- ImageNet Large Scale Visual Recognition ChallengeInternational Journal of Computer Vision, 2015
- Online Regression of Grandmother-Cell Responses with Visual Experience Learning for Face Recognition2014
- Seeing 3D Chairs: Exemplar Part-Based 2D-3D Alignment Using a Large Dataset of CAD Models2014
- Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation2014
- Learning Local Feature Descriptors Using Convex OptimisationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2014
- Transforming Auto-Encoders2011
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009
- Invariant visual representation by single neurons in the human brainNature, 2005
- Learning Optimized Features for Hierarchical Models of Invariant Object RecognitionNeural Computation, 2003