An Attention Model for group-level emotion recognition

TL;DR

This paper introduces an attention-based model that combines global image features and local face features to classify group emotions, achieving competitive results in a challenge.

Contribution

The paper presents a novel attention mechanism to merge global and local face features for improved group emotion recognition.

Findings

Achieved 64.83% accuracy on the EmotiW 2018 test set.

Ensemble of models outperforms individual models.

Ranked 4th in the challenge.

Abstract

In this paper we propose a new approach for classifying the global emotion of images containing groups of people. To achieve this task, we consider two different and complementary sources of information: i) a global representation of the entire image (ii) a local representation where only faces are considered. While the global representation of the image is learned with a convolutional neural network (CNN), the local representation is obtained by merging face features through an attention mechanism. The two representations are first learned independently with two separate CNN branches and then fused through concatenation in order to obtain the final group-emotion classifier. For our submission to the EmotiW 2018 group-level emotion recognition challenge, we combine several variations of the proposed model into an ensemble, obtaining a final accuracy of 64.83% on the test set and ranking 4th among all challenge participants.