Real-time Convolutional Neural Networks for Emotion and Gender Classification

TL;DR

This paper presents a framework for designing real-time convolutional neural networks capable of face detection, gender, and emotion classification simultaneously, validated through benchmark datasets and deployment on a robot.

Contribution

The paper introduces a general CNN building framework for real-time emotion and gender classification, including a novel visualization technique and open-source implementation.

Findings

Achieved 96% accuracy on IMDB gender dataset

Achieved 66% accuracy on FER-2013 emotion dataset

Validated real-time performance on a robotic platform

Abstract

In this paper we propose an implement a general convolutional neural network (CNN) building framework for designing real-time CNNs. We validate our models by creating a real-time vision system which accomplishes the tasks of face detection, gender classification and emotion classification simultaneously in one blended step using our proposed CNN architecture. After presenting the details of the training procedure setup we proceed to evaluate on standard benchmark sets. We report accuracies of 96% in the IMDB gender dataset and 66% in the FER-2013 emotion dataset. Along with this we also introduced the very recent real-time enabled guided back-propagation visualization technique. Guided back-propagation uncovers the dynamics of the weight changes and evaluates the learned features. We argue that the careful implementation of modern CNN architectures, the use of the current regularization methods and the visualization of previously hidden features are necessary in order to reduce the gap between slow performances and real-time architectures. Our system has been validated by its deployment on a Care-O-bot 3 robot used during RoboCup@Home competitions. All our code, demos and pre-trained architectures have been released under an open-source license in our public repository.