Mixture separability loss in a deep convolutional network for image classification. Issue 1 (1st January 2019)

• Record Type: Journal Article
• Authors: Do, Trung Dung; Jin, Cheng‐Bin; Nguyen, Van Huan; Kim, Hakil
• Is Part Of: IET image processing. Volume 13:Issue 1(2019)
• Page Start: 135
• Page End: 141
• Publication Date: 2019-01-01
• Subjects: image representation -- image classification -- entropy -- computer vision -- learning (artificial intelligence) -- feedforward neural nets; intra‐class images -- loss function -- MSL -- deep convolutional network -- image classification -- machine learning -- well‐known networks -- network structures -- applying cross‐entropy loss -- training images -- novel cost function -- between‐class loss -- inter‐class images -- convolutional layers -- mixture separability loss -- within‐class loss -- street view house number -- Canadian institute for advanced research -- self‐collected Inha computer vision lab gender dataset; Image processing -- Periodicals; 621.36705
• DOI: 10.1049/iet-ipr.2018.5613
• Languages: English
• ISSNs: 1751-9659

Abstract

Abstract : In machine learning, the cost function is crucial because it measures how good or bad a system is. In image classification, well‐known networks only consider modifying the network structures and applying cross‐entropy loss at the end of the network. However, using only cross‐entropy loss causes a network to stop updating weights when all training images are correctly classified. This is the problem of early saturation. This study proposes a novel cost function, called mixture separability loss (MSL), which updates the weights of the network even when most of the training images are accurately predicted. MSL consists of between‐class and within‐class loss. Between‐class loss maximises the differences between inter‐class images, whereas within‐class loss minimises the similarities between intra‐class images. They designed the proposed loss function to attach to different convolutional layers in the network in order to utilise intermediate feature maps. Experiments show that a network with MSL deepens the learning process and obtains promising results with some public datasets, such as Street View House Number, Canadian Institute for Advanced Research, and the authors' self‐collected Inha Computer Vision Lab gender dataset.