Performance evaluation of SLAM-ASR: The Good, the Bad, the Ugly, and the Way Forward

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kumar, Shashi, Thorbecke, Iuliia, Burdisso, Sergio, Villatoro-Tello, Esaú, E, Manjunath K, Hacioğlu, Kadri, Rangappa, Pradeep, Motlicek, Petr, Ganapathiraju, Aravind, Stolcke, Andreas
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929683815202816
author Kumar, Shashi
Thorbecke, Iuliia
Burdisso, Sergio
Villatoro-Tello, Esaú
E, Manjunath K
Hacioğlu, Kadri
Rangappa, Pradeep
Motlicek, Petr
Ganapathiraju, Aravind
Stolcke, Andreas
author_facet Kumar, Shashi
Thorbecke, Iuliia
Burdisso, Sergio
Villatoro-Tello, Esaú
E, Manjunath K
Hacioğlu, Kadri
Rangappa, Pradeep
Motlicek, Petr
Ganapathiraju, Aravind
Stolcke, Andreas
contents Recent research has demonstrated that training a linear connector between speech foundation encoders and large language models (LLMs) enables this architecture to achieve strong ASR capabilities. Despite the impressive results, it remains unclear whether these simple approaches are robust enough across different scenarios and speech conditions, such as domain shifts and speech perturbations. In this paper, we address these questions by conducting various ablation experiments using a recent and widely adopted approach called SLAM-ASR. We present novel empirical findings that offer insights on how to effectively utilize the SLAM-ASR architecture across a wide range of settings. Our main findings indicate that SLAM-ASR exhibits poor performance in cross-domain evaluation settings. Additionally, speech perturbations on in-domain data, such as changes in speech rate or additive noise, can significantly degrade performance. Our findings offer critical insights for fine-tuning and configuring robust LLM-based ASR models, tailored to different data characteristics and computational resources.
format Preprint
id arxiv_https___arxiv_org_abs_2411_03866
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Performance evaluation of SLAM-ASR: The Good, the Bad, the Ugly, and the Way Forward
Kumar, Shashi
Thorbecke, Iuliia
Burdisso, Sergio
Villatoro-Tello, Esaú
E, Manjunath K
Hacioğlu, Kadri
Rangappa, Pradeep
Motlicek, Petr
Ganapathiraju, Aravind
Stolcke, Andreas
Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
Recent research has demonstrated that training a linear connector between speech foundation encoders and large language models (LLMs) enables this architecture to achieve strong ASR capabilities. Despite the impressive results, it remains unclear whether these simple approaches are robust enough across different scenarios and speech conditions, such as domain shifts and speech perturbations. In this paper, we address these questions by conducting various ablation experiments using a recent and widely adopted approach called SLAM-ASR. We present novel empirical findings that offer insights on how to effectively utilize the SLAM-ASR architecture across a wide range of settings. Our main findings indicate that SLAM-ASR exhibits poor performance in cross-domain evaluation settings. Additionally, speech perturbations on in-domain data, such as changes in speech rate or additive noise, can significantly degrade performance. Our findings offer critical insights for fine-tuning and configuring robust LLM-based ASR models, tailored to different data characteristics and computational resources.
title Performance evaluation of SLAM-ASR: The Good, the Bad, the Ugly, and the Way Forward
topic Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2411.03866