Dynamic Graph-Guided Transferable Regression for Cross-Domain Speech Emotion Recognition

To deal with the problem of cross-domain speech emotion recognition (SER), in this paper, we propose a novel dynamic graph-guided transferable regression (DGTR) method. Specifically, a retargeted discriminant linear regression in the source domain is utilized to make the projection matrix discriminative. Meanwhile, an adaptive maximum entropy graph is designed for similarity measurement for different domains. Experiments on four popular datasets show that our method can achieve better performance compared with several related state-of-the-art methods.