Saved in:
Bibliographic Details
Main Authors: Yadav, Hemant, Shah, Rajiv Ratn, Sitaram, Sunayana
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2410.11086
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910833790943232
author Yadav, Hemant
Shah, Rajiv Ratn
Sitaram, Sunayana
author_facet Yadav, Hemant
Shah, Rajiv Ratn
Sitaram, Sunayana
contents Information in speech can be categorized into two groups: Content (what is being said, such as linguistics) and Other (how it is expressed such as information about speaker and paralinguistic features). Current self-supervised learning (SSL) methods are shown to divide the model's representational-depth or layers in two, with earlier layers specializing in Other and later layers in Content related tasks. This layer-wise division is inherently sub-optimal, as neither information type can use all layers to build hierarchical representations. To address this, we propose JOOCI, a novel speech representation learning method that does not compromise on the representational-depth for either information type. JOOCI outperforms WavLM by 26.5%, and other models of similar size (100M parameters), when evaluated on two speaker recognition and two language tasks from the SUPERB benchmark, demonstrating its effectiveness in Jointly Optimizing Other and Content Information (JOOCI).
format Preprint
id arxiv_https___arxiv_org_abs_2410_11086
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle JOOCI: a Framework for Learning Comprehensive Speech Representations
Yadav, Hemant
Shah, Rajiv Ratn
Sitaram, Sunayana
Computation and Language
Information in speech can be categorized into two groups: Content (what is being said, such as linguistics) and Other (how it is expressed such as information about speaker and paralinguistic features). Current self-supervised learning (SSL) methods are shown to divide the model's representational-depth or layers in two, with earlier layers specializing in Other and later layers in Content related tasks. This layer-wise division is inherently sub-optimal, as neither information type can use all layers to build hierarchical representations. To address this, we propose JOOCI, a novel speech representation learning method that does not compromise on the representational-depth for either information type. JOOCI outperforms WavLM by 26.5%, and other models of similar size (100M parameters), when evaluated on two speaker recognition and two language tasks from the SUPERB benchmark, demonstrating its effectiveness in Jointly Optimizing Other and Content Information (JOOCI).
title JOOCI: a Framework for Learning Comprehensive Speech Representations
topic Computation and Language
url https://arxiv.org/abs/2410.11086