A study on cross-corpus speech emotion recognition and data augmentation
This addresses robustness issues in speech emotion recognition for applications like human-computer interaction, but it is incremental as it builds on existing methods with new data combinations.
The study tackled the problem of speech emotion recognition models performing poorly on unseen speakers or acoustic conditions by investigating cross-corpus training and data augmentation, finding that mixed corpora reduced performance drops from 10-40% to 1-8% in mismatched conditions and data augmentation added up to 4% gains.
Models that can handle a wide range of speakers and acoustic conditions are essential in speech emotion recognition (SER). Often, these models tend to show mixed results when presented with speakers or acoustic conditions that were not visible during training. This paper investigates the impact of cross-corpus data complementation and data augmentation on the performance of SER models in matched (test-set from same corpus) and mismatched (test-set from different corpus) conditions. Investigations using six emotional speech corpora that include single and multiple speakers as well as variations in emotion style (acted, elicited, natural) and recording conditions are presented. Observations show that, as expected, models trained on single corpora perform best in matched conditions while performance decreases between 10-40% in mismatched conditions, depending on corpus specific features. Models trained on mixed corpora can be more stable in mismatched contexts, and the performance reductions range from 1 to 8% when compared with single corpus models in matched conditions. Data augmentation yields additional gains up to 4% and seem to benefit mismatched conditions more than matched ones.