What is it about?
This paper explains our solutions for the Stuttering and the Vocalisation tasks within the ACM Multimedia 2022 Computational Paralinguistics ChallengE. The goal of this competition is to develop Artificial Intelligence models that can recognize different aspects (stutterings and emotions) of the speaker using only their recorded voice.
Featured Image
Photo by Icons8 Team on Unsplash
Why is it important?
Our system achieved excellent results in both competitions and showcases how we can recognize different kinds of stutterings (e.g. word/phrase repetition, prolongation, sound repetition, etc.) and emotions in non-verbal vocal expressions (such as laughter, cries, moans, and screams). In both challenges, our solution achieved the best results, demonstrating their effectiveness.
Perspectives
Participating in these competitions is quite important for me. On the one hand, it poses new challenges for me and often makes me think "outside of the box", to come up with novel solutions. At the same time, it allows us to compete with other groups in a controlled setup, using the same data and under the same limitations.
Tamas Grosz
Aalto-yliopisto
Read the Original
This page is a summary of: Wav2vec2-based Paralinguistic Systems to Recognise Vocalised Emotions and Stuttering, October 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3503161.3551572.
You can read the full text:
Contributors
The following have contributed to this page
