Unsupervised low-rank representations for speech emotion recognition
This work addresses the curse of dimensionality for researchers and practitioners in speech emotion recognition, but it is incremental as it applies existing dimensionality reduction techniques to this domain.
The paper tackled the problem of high-dimensional feature sets in speech emotion recognition by applying linear and non-linear dimensionality reduction algorithms, resulting in performance improvements across various classification methods and databases.
We examine the use of linear and non-linear dimensionality reduction algorithms for extracting low-rank feature representations for speech emotion recognition. Two feature sets are used, one based on low-level descriptors and their aggregations (IS10) and one modeling recurrence dynamics of speech (RQA), as well as their fusion. We report speech emotion recognition (SER) results for learned representations on two databases using different classification methods. Classification with low-dimensional representations yields performance improvement in a variety of settings. This indicates that dimensionality reduction is an effective way to combat the curse of dimensionality for SER. Visualization of features in two dimensions provides insight into discriminatory abilities of reduced feature sets.