SER Evals: In-domain and Out-of-domain Benchmarking for Speech Emotion Recognition

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Osman, Mohamed, Kaplan, Daniel Z., Nadeem, Tamer
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914913325154304
author Osman, Mohamed
Kaplan, Daniel Z.
Nadeem, Tamer
author_facet Osman, Mohamed
Kaplan, Daniel Z.
Nadeem, Tamer
contents Speech emotion recognition (SER) has made significant strides with the advent of powerful self-supervised learning (SSL) models. However, the generalization of these models to diverse languages and emotional expressions remains a challenge. We propose a large-scale benchmark to evaluate the robustness and adaptability of state-of-the-art SER models in both in-domain and out-of-domain settings. Our benchmark includes a diverse set of multilingual datasets, focusing on less commonly used corpora to assess generalization to new data. We employ logit adjustment to account for varying class distributions and establish a single dataset cluster for systematic evaluation. Surprisingly, we find that the Whisper model, primarily designed for automatic speech recognition, outperforms dedicated SSL models in cross-lingual SER. Our results highlight the need for more robust and generalizable SER models, and our benchmark serves as a valuable resource to drive future research in this direction.
format Preprint
id arxiv_https___arxiv_org_abs_2408_07851
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SER Evals: In-domain and Out-of-domain Benchmarking for Speech Emotion Recognition
Osman, Mohamed
Kaplan, Daniel Z.
Nadeem, Tamer
Computation and Language
Artificial Intelligence
Speech emotion recognition (SER) has made significant strides with the advent of powerful self-supervised learning (SSL) models. However, the generalization of these models to diverse languages and emotional expressions remains a challenge. We propose a large-scale benchmark to evaluate the robustness and adaptability of state-of-the-art SER models in both in-domain and out-of-domain settings. Our benchmark includes a diverse set of multilingual datasets, focusing on less commonly used corpora to assess generalization to new data. We employ logit adjustment to account for varying class distributions and establish a single dataset cluster for systematic evaluation. Surprisingly, we find that the Whisper model, primarily designed for automatic speech recognition, outperforms dedicated SSL models in cross-lingual SER. Our results highlight the need for more robust and generalizable SER models, and our benchmark serves as a valuable resource to drive future research in this direction.
title SER Evals: In-domain and Out-of-domain Benchmarking for Speech Emotion Recognition
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2408.07851