Boosting Mapping Functionality of Neural Networks via Latent Feature Generation based on Reversible Learning

TL;DR

This paper introduces reversible learning to generate latent features for hard samples, enhancing neural network mapping in visual recognition without extra data augmentation, and demonstrates superior performance on multiple datasets.

Contribution

Proposes a novel reversible learning method for generating latent features to improve neural network mapping in visual recognition tasks.

Findings

Outperforms state-of-the-art methods on MNIST, Cifar-10/100, and EBPC datasets.

Effectively improves network mapping capability without additional data augmentation.

Demonstrates efficiency in handling biased and poorly categorized datasets.

Abstract

This paper addresses a boosting method for mapping functionality of neural networks in visual recognition such as image classification and face recognition. We present reversible learning for generating and learning latent features using the network itself. By generating latent features corresponding to hard samples and applying the generated features in a training stage, reversible learning can improve a mapping functionality without additional data augmentation or handling the bias of dataset. We demonstrate an efficiency of the proposed method on the MNIST,Cifar-10/100, and Extremely Biased and poorly categorized dataset (EBPC dataset). The experimental results show that the proposed method can outperform existing state-of-the-art methods in visual recognition. Extensive analysis shows that our method can efficiently improve the mapping capability of a network.