Visual analytics system for convolutional neural network based classifiers

Abstract

A visual analytics method and system is disclosed for visualizing an operation of an image classification model having at least one convolutional neural network layer. The image classification model classifies sample images into one of a predefined set of possible classes. The visual analytics method determines a unified ordering of the predefined set of possible classes based on a similarity hierarchy such that classes that are similar to one another are clustered together in the unified ordering. The visual analytics method displays various graphical depictions, including a class hierarchy viewer, a confusion matrix, and a response map. In each case, the elements of the graphical depictions are arranged in accordance with the unified ordering. Using the method, a user a better able to understand the training process of the model, diagnose the separation power of the different feature detectors of the model, and improve the architecture of the model.

Description

[0001] This application claims the benefit of priority to U.S. Provisional Application Serial No. 62 / 537,613, filed July 27, 2017, the disclosure of which is hereby incorporated by reference in its entirety. technical field [0002] The devices and methods disclosed in this document relate to convolutional neural networks, and more particularly, to visual analytics for convolutional neural networks. Background technique [0003] Unless otherwise indicated herein, the materials described in this section are not prior art to the claims in this application and are not admitted to be prior art by inclusion in this section. [0004] Object recognition is a fundamental problem in computer vision that involves classifying images into a predefined number of classes. Convolutional neural networks (CNNs) have achieved state-of-the-art results on this problem, thanks to the availability of large and labeled datasets and powerful computing infrastructure. CNNs automatically extract dis...

AI Technical Summary

Problems solved by technology

However, little attention has been paid to visualizing the classification error itself and refining the CNN accordingly

Images