Difference between revisions of "Projects:2020s1-1410 Speech Enhancement for Automatic Speech Recognition"
m (Used skeleton template) |
(Filled in template a little, should be enough for the time being.) |
||
| Line 2: | Line 2: | ||
[[Category:Final Year Projects]] | [[Category:Final Year Projects]] | ||
[[Category:2020s1|1410]] | [[Category:2020s1|1410]] | ||
| − | + | Speech recognition is becoming more and more widely used, though the input audio to these systems is rarely clean. A number of techniques | |
| + | <ref name="SEGAN">Pascual, S., Bonafonte, A. and Serra, J., 2017. ''SEGAN: Speech enhancement generative adversarial network''. arXiv preprint arXiv:1703.09452</ref> | ||
| + | <ref name="Wave-U-Net">Stoller, D., Ewert, S. and Dixon, S., 2018. ''Wave-u-net: A multi-scale neural network for end-to-end audio source separation''. arXiv preprint arXiv:1806.03185</ref> | ||
| + | have been developed to reduce the background noise of speech clips, both using deep neural networks, and more traditional filters. | ||
| + | |||
| + | The overall objective of this project is to compare a number of speech enhancement techniques in a fair environment, and to also compare the results of each technique after its output is fed through an automatic speech recogniser. | ||
| + | |||
== Introduction == | == Introduction == | ||
| − | This project | + | ''This project follows from work done previously by University of Adelaide students Jordan Parker, Shalin Shah, and Nha Nam (Harry) Nguyen as a summer scholarship project.'' |
=== Project team === | === Project team === | ||
| Line 18: | Line 24: | ||
=== Objectives === | === Objectives === | ||
| − | + | ==== Obtain a dataset ==== | |
| + | |||
| + | Each speech enhancement method has been demonstrated by using different audio datasets depending on the creator(s). Despite this, the general concept is very similar: | ||
| + | * Collect a large amount of "noise" audio | ||
| + | * Collect a large amount of clean speech audio - if a transcription exists too, this is called a '''corpus''' | ||
| + | * Combine the two datasets to synthesise noisy speech audio | ||
| + | |||
| + | The goal for this objective is to generate the means of creating a very large (approx 1000hrs) dataset of mixed audio, while maintaining a record of the original clean and noise files - as some methods use these during training. This dataset / generation methodology can then be used by all methods for a fair comparison. | ||
| + | |||
| + | ==== Train and optimise ==== | ||
| + | |||
| + | A number of promising techniques are selected, and their models trained on the dataset from the previous objective. For non-learning methods, their algorithms may be optimised or altered in some small manner to generate the best results. | ||
| + | |||
| + | ==== Compare methods ==== | ||
| + | |||
| + | |||
== Background == | == Background == | ||
| Line 30: | Line 51: | ||
== References == | == References == | ||
| − | + | <references /> | |
| − | |||
| − | |||
Revision as of 10:53, 21 April 2020
Speech recognition is becoming more and more widely used, though the input audio to these systems is rarely clean. A number of techniques [1] [2] have been developed to reduce the background noise of speech clips, both using deep neural networks, and more traditional filters.
The overall objective of this project is to compare a number of speech enhancement techniques in a fair environment, and to also compare the results of each technique after its output is fed through an automatic speech recogniser.
Contents
Introduction
This project follows from work done previously by University of Adelaide students Jordan Parker, Shalin Shah, and Nha Nam (Harry) Nguyen as a summer scholarship project.
Project team
Project students
- Patrick Gregory
- Zachary Knopoff
Supervisors
- Dr. Said Al-Sarawi
- Dr. Ahmad Hashemi-Sakhtsari (DST Group)
- Mr. Paul Jager (DST Group)
Advisors
- Ms. Wei Gao (Emily)
Objectives
Obtain a dataset
Each speech enhancement method has been demonstrated by using different audio datasets depending on the creator(s). Despite this, the general concept is very similar:
- Collect a large amount of "noise" audio
- Collect a large amount of clean speech audio - if a transcription exists too, this is called a corpus
- Combine the two datasets to synthesise noisy speech audio
The goal for this objective is to generate the means of creating a very large (approx 1000hrs) dataset of mixed audio, while maintaining a record of the original clean and noise files - as some methods use these during training. This dataset / generation methodology can then be used by all methods for a fair comparison.
Train and optimise
A number of promising techniques are selected, and their models trained on the dataset from the previous objective. For non-learning methods, their algorithms may be optimised or altered in some small manner to generate the best results.
