Emotion recognition based on feedback weighted fusion of multimodal emotion data

Emotion recognition with decision-level weighted fusion based on multimodal emotion data is a challenging research topic and has attracted great attention in the past few years. This paper presents a novel method, utilizing recognition rate of single set of emotion feature to calculate weight. First, we concatenate a physiological feature and facial expression feature together form two sets of multimodal emotion feature. Then, we calculate each emotion recognition rate of each sets of emotion feature and get weight matrix. Lastly, we construct a classifier based on decision-level weighted fusion strategy. And the experiments on the MAHNOB-HCI database show that our method has achieved encouraging recognition results compared to the state-of-the-art methods.