Learning Affective Features With a Hybrid Deep Model for Audio–Visual Emotion Recognition 张石清, ZhangShiliang, HuangTiejun, GaoWen, TianQi 10月 1, 2018 DOI