Feasibility study of spiking neural network for voice classification
Spiking neural network (SNN), largely inspired by a biological neural network, offers many advantages over the conventional deep neural network (DNN) [1]/ convolutional neural network (CNN) [2] in energy efficiency and latency due to its event based processing, which makes it a good candidate...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2019
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/78456 |
| _version_ | 1811684257996734464 |
|---|---|
| author | Zhang, Zhongyi |
| author2 | Goh Wang Ling |
| author_facet | Goh Wang Ling Zhang, Zhongyi |
| author_sort | Zhang, Zhongyi |
| collection | NTU |
| description | Spiking neural network (SNN), largely inspired by a biological neural network, offers
many advantages over the conventional deep neural network (DNN) [1]/ convolutional
neural network (CNN) [2] in energy efficiency and latency due to its event based
processing, which makes it a good candidate for energy efficient voice classification.
Furthermore, the learning mechanism of the SNN typically requires only local
information of pre-synaptic neuron and post-synaptic neuron when a spike happens,
providing a light-weighted energy-efficient and hardware-friendly solution for the
applications of voice recognition and classification.
This dissertation reports a digital spiking neuron based on Leaky Integrate-and-Fire
(LIF) model [3]. As the basic units of neurons in SNN, several LIF models jointly
construct a 3-layer neural network. Subsequently, eight Infinite Impulse Response (IIR)
digital filters ranging from 0 Hz to 400 Hz are designed to analyze human voice in both
time and frequency domains, which extracts the feature of male and female voices. Two
thousands male and female voice clips are used as training sets and five hundred voices
are used as test sets in the neural network. The functionality and performance of the
proposed digital spiking neuron can be verified by test sets to recognize male and
female voice. The obtained results and simulations in MATLAB demonstrate the
superiority of the proposed SNN and determine the potential of such systems in voice
classification. |
| first_indexed | 2024-10-01T04:25:46Z |
| format | Thesis |
| id | ntu-10356/78456 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T04:25:46Z |
| publishDate | 2019 |
| record_format | dspace |
| spelling | ntu-10356/784562023-07-04T16:20:48Z Feasibility study of spiking neural network for voice classification Zhang, Zhongyi Goh Wang Ling School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering Spiking neural network (SNN), largely inspired by a biological neural network, offers many advantages over the conventional deep neural network (DNN) [1]/ convolutional neural network (CNN) [2] in energy efficiency and latency due to its event based processing, which makes it a good candidate for energy efficient voice classification. Furthermore, the learning mechanism of the SNN typically requires only local information of pre-synaptic neuron and post-synaptic neuron when a spike happens, providing a light-weighted energy-efficient and hardware-friendly solution for the applications of voice recognition and classification. This dissertation reports a digital spiking neuron based on Leaky Integrate-and-Fire (LIF) model [3]. As the basic units of neurons in SNN, several LIF models jointly construct a 3-layer neural network. Subsequently, eight Infinite Impulse Response (IIR) digital filters ranging from 0 Hz to 400 Hz are designed to analyze human voice in both time and frequency domains, which extracts the feature of male and female voices. Two thousands male and female voice clips are used as training sets and five hundred voices are used as test sets in the neural network. The functionality and performance of the proposed digital spiking neuron can be verified by test sets to recognize male and female voice. The obtained results and simulations in MATLAB demonstrate the superiority of the proposed SNN and determine the potential of such systems in voice classification. Master of Science (Electronics) 2019-06-20T04:51:03Z 2019-06-20T04:51:03Z 2019 Thesis http://hdl.handle.net/10356/78456 en 79 p. application/pdf |
| spellingShingle | DRNTU::Engineering::Electrical and electronic engineering Zhang, Zhongyi Feasibility study of spiking neural network for voice classification |
| title | Feasibility study of spiking neural network for voice classification |
| title_full | Feasibility study of spiking neural network for voice classification |
| title_fullStr | Feasibility study of spiking neural network for voice classification |
| title_full_unstemmed | Feasibility study of spiking neural network for voice classification |
| title_short | Feasibility study of spiking neural network for voice classification |
| title_sort | feasibility study of spiking neural network for voice classification |
| topic | DRNTU::Engineering::Electrical and electronic engineering |
| url | http://hdl.handle.net/10356/78456 |
| work_keys_str_mv | AT zhangzhongyi feasibilitystudyofspikingneuralnetworkforvoiceclassification |