MELT: Towards Automated Multimodal Emotion Data Annotation by Leveraging LLM Embedded Knowledge

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Jing, Xin, Wang, Jiadong, Tsangko, Iosif, Triantafyllopoulos, Andreas, Schuller, Björn W.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908386028683264
author Jing, Xin
Wang, Jiadong
Tsangko, Iosif
Triantafyllopoulos, Andreas
Schuller, Björn W.
author_facet Jing, Xin
Wang, Jiadong
Tsangko, Iosif
Triantafyllopoulos, Andreas
Schuller, Björn W.
contents Although speech emotion recognition (SER) has advanced significantly with deep learning, annotation remains a major hurdle. Human annotation is not only costly but also subject to inconsistencies annotators often have different preferences and may lack the necessary contextual knowledge, which can lead to varied and inaccurate labels. Meanwhile, Large Language Models (LLMs) have emerged as a scalable alternative for annotating text data. However, the potential of LLMs to perform emotional speech data annotation without human supervision has yet to be thoroughly investigated. To address these problems, we apply GPT-4o to annotate a multimodal dataset collected from the sitcom Friends, using only textual cues as inputs. By crafting structured text prompts, our methodology capitalizes on the knowledge GPT-4o has accumulated during its training, showcasing that it can generate accurate and contextually relevant annotations without direct access to multimodal inputs. Therefore, we propose MELT, a multimodal emotion dataset fully annotated by GPT-4o. We demonstrate the effectiveness of MELT by fine-tuning four self-supervised learning (SSL) backbones and assessing speech emotion recognition performance across emotion datasets. Additionally, our subjective experiments\' results demonstrate a consistence performance improvement on SER.
format Preprint
id arxiv_https___arxiv_org_abs_2505_24493
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MELT: Towards Automated Multimodal Emotion Data Annotation by Leveraging LLM Embedded Knowledge
Jing, Xin
Wang, Jiadong
Tsangko, Iosif
Triantafyllopoulos, Andreas
Schuller, Björn W.
Artificial Intelligence
Sound
Audio and Speech Processing
Although speech emotion recognition (SER) has advanced significantly with deep learning, annotation remains a major hurdle. Human annotation is not only costly but also subject to inconsistencies annotators often have different preferences and may lack the necessary contextual knowledge, which can lead to varied and inaccurate labels. Meanwhile, Large Language Models (LLMs) have emerged as a scalable alternative for annotating text data. However, the potential of LLMs to perform emotional speech data annotation without human supervision has yet to be thoroughly investigated. To address these problems, we apply GPT-4o to annotate a multimodal dataset collected from the sitcom Friends, using only textual cues as inputs. By crafting structured text prompts, our methodology capitalizes on the knowledge GPT-4o has accumulated during its training, showcasing that it can generate accurate and contextually relevant annotations without direct access to multimodal inputs. Therefore, we propose MELT, a multimodal emotion dataset fully annotated by GPT-4o. We demonstrate the effectiveness of MELT by fine-tuning four self-supervised learning (SSL) backbones and assessing speech emotion recognition performance across emotion datasets. Additionally, our subjective experiments\' results demonstrate a consistence performance improvement on SER.
title MELT: Towards Automated Multimodal Emotion Data Annotation by Leveraging LLM Embedded Knowledge
topic Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2505.24493