Estimating the Uncertainty in Emotion Class Labels with Utterance-Specific Dirichlet Priors

Abstract

Emotion recognition is a key attribute for artificial intelligence systems that need to naturally interact with humans. However, the task definition is still an open problem due to inherent ambiguity of emotions. In this paper, a novel Bayesian training loss based on per-utterance Dirichlet prior distributions is proposed for verbal emotion recognition, which models the uncertainty in one-hot labels created when human annotators assign the same utterance to different emotion classes. An additional metric is used to evaluate the performance by detecting test utterances with high labelling uncertainty. This removes a major limitation that emotion classification systems only consider utterances with majority labels. Furthermore, a frequentist approach is studied to leverage the continuous-valued "soft" labels obtained by averaging the one-hot labels. We propose a two-branch model structure for emotion classification on a per-utterance basis. Experiments with the widely used IEMOCAP dataset demonstrate that the two-branch structure achieves state-of-the-art classification results with all common IEMOCAP test setups. Based on this, uncertainty estimation experiments were performed. The best performance in terms of the area under the precision-recall curve when detecting utterances with high uncertainty was achieved by interpolating the Bayesian training loss with the Kullback-Leibler divergence training loss for the soft labels.