کنفرانس مهندسی برق ایران

صفحه اصلی / سی و یکمین کنفرانس بین المللی مهندسی برق

Speech Emotion Recognition Using Transfer Learning and Self-Supervised Speech Representation Learning

نویسندگان :

Marziye Azad¹ Babak Nasersharif²

1- دانشگاه صنعتی خواجه نصیرالدین طوسی 2- دانشگاه صنعتی خواجه نصیرالدین طوسی

کلمات کلیدی :

HuBERT،Conformer،self-supervised learning،speech emotion recognition

چکیده :

Self-supervised speech representation learning (S3RL) models like wav2vec2.0, Hidden-unit BERT (HuBERT), and WavLM are trained with a great amount of speech data and subsequently give a general purpose speech representation that then needs to be finetuned for different speech processing tasks like ASR. Despite these models’ good performance, they suffer from massive structures and a great number of parameters which makes their finetuning inapplicable for low-resource tasks like speech emotion recognition. In this paper, a small model is proposed for speech emotion recognition based on the Hubert model by transferring the Hubert convolutional feature encoder and substituting all of its transformers with a simple conformer block. Then this simple model is trained with emotional speech signals. The experimental results demonstrate that the proposed model has comparable results with other well-performing S3RL models.

لیست مقالات

لیست مقالات بایگانی شده

Solving the inverse problem for EEG signals when learning a new motor task using GRU neural network

Milad Khosravi - Fariba Bahrami - Behzad Moshiri - Ahmad Kalhor

Anomaly Detection in Urban Water Distribution Grids Using Fog Computing Architecture

Sara Mirzaie - Mohammadreza Avazaghaei - Omid Bushehrian

Ultra-broadband and compact beamsplitters using subwavelength-grating-assisted zero gap directional couplers

Kamalodin Arik - Mahmood Akbari - Amin Khavasi

Photonic Crystal-based Plasmonic Biosensor with Low-cost and High-sensitivity Properties

Mahdieh Ahmadi Motlagh - Mahdieh Bozorgi - Mahmood Rafaei-Booket

Enhancing the Incident Angle Band in Carpet Cloaking using Deep Neural Networks

Amirhossein Fallah - Leila Yousefi - Ahmad Kalhor

The Conduction Mechanism in Micron-Thick ZnO Layers Grown on Si Substrates by Spray Pyrolysis

Mohsen Gharesi - Alireza Karimpour - Reza Razmand - Faramarz Hossein-Babaei

High-Gain Quasi-Z-Source DC-DC Converter with Single Magnetic Core and Pole Placement Control for DC Microgrid Applications

Ali Nadermohammadi - Zahra Behboudi - Amirhossein Akhbari - Soheil Norouzi - Seyed Hossein Hosseini - Mehran Sabahi

Designing Of Type-2 Fuzzy Formation Controller For A Class Of Nonlinear Multiagent System Using JAYA Algorithm

Arvin Attar - Mohammad Ali Badamchizadeh - Sehraneh Ghaemi

تخمین کانال های پهپاد به پهپاد با استفاده از فیلتر کالمن توسعه یافته

فهیمه رنجبر - محمدعلی سبقتی

A high speed method for features extraction in face recognition systems

Hosein Khorami - Hadishahriar Shahhoseini

بیشتر

ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 42.0.4