Role of beamforming techniques in the future for IoT, artificial intelligence, and real time processing
Speech recognition from a distance, also known as far-field automatic speech recognition, uses machine learning for processing. However, environmental conditions often corrupt speech recorded from a distance, causing disturbances. To obtain desired speech from corrupted signals, various techniq...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2024
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/110983/1/Role%20of%20beamforming%20techniques%20in%20the%20future%20for%20IoT%2C%20Artificial%20Intelligence%20and%20Real%20Time%20Processing%20Final%20copy%2026-10-23.pdf http://psasir.upm.edu.my/id/eprint/110983/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Putra Malaysia |
| Language: | English |
| Summary: | Speech recognition from a distance, also known
as far-field automatic speech recognition, uses machine learning
for processing. However, environmental conditions often
corrupt speech recorded from a distance, causing disturbances.
To obtain desired speech from corrupted signals, various
techniques are used, such as de-reverberation, source
separation, denoising, and acoustic beamforming. The aim is to
design a robust and multi-condition adaptive system in far-fieldbased automatic speech recognition systems. This review paper
focuses on speech enhancement for the future of speech with
progressive technologies like deep learning and machine
learning. It highlights the extensive research on beamformingbased speech enhancement over the past few years, based on
different techniques, performance, advantages, limitations, and
scope for improvement. Finally, this paper explores the smart
city applications that benefited from speech enhancement and
beamforming. |
|---|