Disentanglement for audio-visual emotion recognition using multitask setup
This work addresses the issue of entangled representations in multitask learning for emotion recognition, which is incremental as it builds on existing multitask approaches to improve feature separation.
The paper tackled the problem of entangled information in multitask models for audio-visual emotion recognition by developing a framework to disentangle emotion-specific features from person identity, achieving up to 13% disentanglement while preserving emotion recognition performance.
Deep learning models trained on audio-visual data have been successfully used to achieve state-of-the-art performance for emotion recognition. In particular, models trained with multitask learning have shown additional performance improvements. However, such multitask models entangle information between the tasks, encoding the mutual dependencies present in label distributions in the real world data used for training. This work explores the disentanglement of multimodal signal representations for the primary task of emotion recognition and a secondary person identification task. In particular, we developed a multitask framework to extract low-dimensional embeddings that aim to capture emotion specific information, while containing minimal information related to person identity. We evaluate three different techniques for disentanglement and report results of up to 13% disentanglement while maintaining emotion recognition performance.