Global-local attention for emotion recognition

Abstract Human emotion recognition is an active research area in artificial intelligence and has made substantial progress over the past few years. Many recent works mainly focus on facial regions to infer human affection, while surrounding context information not effectively utilized. In this paper, we proposed a new deep network recognize emotions using novel global-local attention mechanism. Our designed extract features from both independently, then learn them together module. way, contextual used emotions, therefore enhancing discrimination of classifier. The intensive experiments show that our method surpasses current state-of-the-art methods datasets by fair margin. Qualitatively, module can more meaningful maps than previous methods. source code trained model are available at https://github.com/minhnhatvt/glamor-net .