Exploring the Deep Feature Space of a Cell Classification Neural Network

TL;DR

This paper visualizes and analyzes the feature space of a deep neural network trained for white blood cell classification, revealing class separation, feature development, and learned features through advanced visualization techniques.

Contribution

It introduces and applies visualization methods to interpret the feature space of a cell classification neural network, enhancing understanding of learned representations.

Findings

Class separation and structure visualized at various network depths

Development of feature complexity as network depth increases

Identification of salient features using modulated feature images

Abstract

In this paper, we present contemporary techniques for visualising the feature space of a deep learning image classification neural network. These techniques are viewed in the context of a feed-forward network trained to classify low resolution fluorescence images of white blood cells captured using optofluidic imaging. The model has two output classes corresponding to two different cell types, which are often difficult to distinguish by eye. This paper has two major sections. The first looks to develop the information space presented by dimension reduction techniques, such as t-SNE, used to embed high-dimensional pre-softmax layer activations into a two-dimensional plane. The second section looks at feature visualisation by optimisation to generate feature images representing the learned features of the network. Using and developing these techniques we visualise class separation and structures within the dataset at various depths using clustering algorithms and feature images; track the development of feature complexity as we ascend the network; and begin to extract the features the network has learnt by modulating single-channel feature images with up-scaled neuron activation maps to distinguish their most salient parts.