paper-with-me

홈 › Papers

DistilHuBERT: Speech Representation Learning by Layer-wise Distillation of Hidden-unit BERT

2021-10-05 · Heng-Jui Chang, Shu-wen Yang, Hung-Yi Lee

Self-supervised speech representation learning methods like wav2vec 2.0 and Hidden-unit BERT (HuBERT) leverage unlabeled speech data for pre-training and offer good representations for numerous speech processing tasks. Despite the success of these methods, they require large memory and high pre-training costs, making them inaccessible for researchers in academia and small companies. Therefore, this paper introduces DistilHuBERT, a novel multi-task learning framework to distill hidden representations from a HuBERT model directly. This method reduces HuBERT's size by 75% and 73% faster while retaining most performance in ten different tasks. Moreover, DistilHuBERT required little training time and data, opening the possibilities of pre-training personal and on-device SSL models for speech.

📄 PDF Abstract BibTeX arXiv:2110.01900

Code (1)

s3prl/s3prl/tree/master/s3prl/upstream/distiller 공식 구현 pytorch

Tasks

Multi-Task LearningRepresentation LearningSpeech Representation Learning

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
WordPiece 설명 없음
Adam 설명 없음
Attention Dropout Attention Dropout is a type of dropout used in attention-based architectures, where elements are randomly dropped out of the…
Residual Connection 설명 없음
Weight Decay 설명 없음
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…

Similar Papers 제목 키워드 기반

Improving the Robustness of DistilHuBERT to Unseen Noisy Conditions via Data Augmentation, Curriculum Learning, and Multi-Task Enhancement

2022-11-12 · Heitor R. Guimarães, Arthur Pimentel, Anderson R. Avila, Mehdi Rezagholizadeh 외

Self-supervised speech representation learning aims to extract meaningful factors from the speech signal that can later be used across different downstream tasks, such as speech and/or emotion recognition. Existing model…

Data AugmentationEmotion RecognitionMulti-Task LearningRepresentation Learning+1

Distilling HuBERT with LSTMs via Decoupled Knowledge Distillation

2023-09-18 · Danilo de Oliveira, Timo Gerkmann

Much research effort is being applied to the task of compressing the knowledge of self-supervised models, which are powerful, yet large and memory consuming. In this work, we show that the original method of knowledge di…

Automatic Speech RecognitionKnowledge Distillationspeech-recognitionSpeech Recognition

Emotion Detection in Speech Using Lightweight and Transformer-Based Models: A Comparative and Ablation Study

2025-11-01 · Lucky Onyekwelu-Udoka, Md Shafiqul Islam, Md Shahedul Hasan arxiv

Emotion recognition from speech plays a vital role in the development of empathetic human-computer interaction systems. This paper presents a comparative analysis of lightweight transformer-based models, DistilHuBERT and…

Speech Emotion Recognition

Analyzing Robustness of End-to-End Neural Models for Automatic Speech Recognition

2022-08-17 · Goutham Rajendran, Wei Zou

We investigate robustness properties of pre-trained neural models for automatic speech recognition. Real life data in machine learning is usually very noisy and almost never clean, which can be attributed to various fact…

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)speech-recognitionSpeech Recognition

Ensemble knowledge distillation of self-supervised speech models

2023-02-24 · Kuan-Po Huang, Tzu-hsun Feng, Yu-Kuan Fu, Tsu-Yuan Hsu 외

Distilled self-supervised models have shown competitive performance and efficiency in recent years. However, there is a lack of experience in jointly distilling multiple self-supervised speech models. In our work, we per…

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)Emotion RecognitionKnowledge Distillation+4