The journal IEEE Transactions on Affective Computing (T-AFFC) compiled a selection of five Best Papers published in the journal in 2021. The paper "Improving Cross-Corpus Speech Emotion Recognition with Adversarial Discriminative Doman Generalization" published by Prechter Program researchers Dr. Melvin McInnis and Dr. Emily Mower Provost was selected out of 82 papers for the Best of IEEE Transactions on Affective Computing 2021 Paper Collection.
The awards for the Best of T-AFFC will be presented at the hybrid 10th International Conference on Affective Computing & Intelligent Interaction that will be held online and in Nara, Japan.
Dr. Emily Mower Provost tells us more about the paper:
When our environment changes, the acoustics of our emotion expressions change as well. However, as humans, we know that this does not mean that in a new environment we lose the ability to recognize the emotion expressions of others. Yet, this is often the case for our classifiers. In our paper, our goal was to create methods that identify the commonalities between emotion expressed in different environments and to use these commonalities to recognize emotion in a more generalizable manner. We advocated for the creation of new “meet in the middle” approaches that learn to identify commonalities between emotion expressions from different datasets. We referred to this method as Adversarial Discriminative Domain Generalization (ADDoG). It was inspired by domain generalization techniques, but reflected on a common failing in these approaches applied to emotion recognition: instability and mode collapse. We created a new technique that iteratively moved representations learned for each dataset closer to one another, improving cross-dataset generalization and significantly improving cross-dataset performance.