Hardware-friendly neural network design and optimization for low power IOT applications

As speech becomes a popular way for human to interact with electronic devices in recent years, it leads to interests to apply machine learning in speech related applications, such as sound classification, speech recognition and so on. One of the exciting applications is to develop keyword spottin...

Full description

Saved in:
Bibliographic Details
Main Author: Wang, Yingfeng
Other Authors: Goh Wang Ling
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2022
Subjects:
Online Access:https://hdl.handle.net/10356/157996
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-157996
record_format dspace
spelling sg-ntu-dr.10356-1579962023-07-07T19:13:34Z Hardware-friendly neural network design and optimization for low power IOT applications Wang, Yingfeng Goh Wang Ling School of Electrical and Electronic Engineering A*STAR Institute of Microelectronics Do Anh Tuan EWLGOH@ntu.edu.sg, doat@ime.a-star.edu.sg Engineering::Electrical and electronic engineering As speech becomes a popular way for human to interact with electronic devices in recent years, it leads to interests to apply machine learning in speech related applications, such as sound classification, speech recognition and so on. One of the exciting applications is to develop keyword spotting (KWS) module using neural network. The KWS module is acting as a switch to activate a downstream system, for example, a speech recognition system after certain keywords have been detected. In actual application, a high accuracy KWS which is able to identify the keyword with the existence of background noises is desired for a smooth user experience. Thus, this project aims to design a hardware-friendly and noise-robust neural network for KWS, expecting to classify 10 keywords along with “silence” and “unknown” class. A final LSTM model with 4-bit quantization and k=9 pruning shows 91.74% accuracy on clean audio, with a model size of 7.9KB. Compared to other state-of- the-art KWS architectures classifying the same number of keywords, this work is able to achieve a 2-4% higher accuracy for both clean and noisy audios, as well as a size reduction of at least 29%. Bachelor of Engineering (Electrical and Electronic Engineering) 2022-05-26T12:53:16Z 2022-05-26T12:53:16Z 2022 Final Year Project (FYP) Wang, Y. (2022). Hardware-friendly neural network design and optimization for low power IOT applications. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/157996 https://hdl.handle.net/10356/157996 en B2053-211 application/pdf Nanyang Technological University
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Electrical and electronic engineering
spellingShingle Engineering::Electrical and electronic engineering
Wang, Yingfeng
Hardware-friendly neural network design and optimization for low power IOT applications
description As speech becomes a popular way for human to interact with electronic devices in recent years, it leads to interests to apply machine learning in speech related applications, such as sound classification, speech recognition and so on. One of the exciting applications is to develop keyword spotting (KWS) module using neural network. The KWS module is acting as a switch to activate a downstream system, for example, a speech recognition system after certain keywords have been detected. In actual application, a high accuracy KWS which is able to identify the keyword with the existence of background noises is desired for a smooth user experience. Thus, this project aims to design a hardware-friendly and noise-robust neural network for KWS, expecting to classify 10 keywords along with “silence” and “unknown” class. A final LSTM model with 4-bit quantization and k=9 pruning shows 91.74% accuracy on clean audio, with a model size of 7.9KB. Compared to other state-of- the-art KWS architectures classifying the same number of keywords, this work is able to achieve a 2-4% higher accuracy for both clean and noisy audios, as well as a size reduction of at least 29%.
author2 Goh Wang Ling
author_facet Goh Wang Ling
Wang, Yingfeng
format Final Year Project
author Wang, Yingfeng
author_sort Wang, Yingfeng
title Hardware-friendly neural network design and optimization for low power IOT applications
title_short Hardware-friendly neural network design and optimization for low power IOT applications
title_full Hardware-friendly neural network design and optimization for low power IOT applications
title_fullStr Hardware-friendly neural network design and optimization for low power IOT applications
title_full_unstemmed Hardware-friendly neural network design and optimization for low power IOT applications
title_sort hardware-friendly neural network design and optimization for low power iot applications
publisher Nanyang Technological University
publishDate 2022
url https://hdl.handle.net/10356/157996
_version_ 1772828472765317120