COMPARATIVE ANALYSIS OF BASIC DENOISING METHODS FOR SPEECH IMPROVEMENT USING STATISTICAL MEASURES
Faculty Mentor
Indika Ilandari Dewage
Presentation Type
Poster
Start Date
4-14-2026 9:00 AM
End Date
4-14-2026 11:00 AM
Location
PUB NCR
Primary Discipline of Presentation
Mathematics
Abstract
This research presents a statistical analysis on the effectiveness of four audio denoising techniques: Spectral Gating, Butterworth Filtering, Gaussian Filtering, and Wavelet Denoising. We utilize the signal-to-noise ratio of speech signals (SNR), Root Mean Square Error (RMSE), and Short-Time Objective Intelligibility (STOI) as statistical metrics to compare the effectiveness of the denoising techniques. After statistical comparison we establish that the Spectral Gating denoising technique is superior to the other techniques in denoising and producing an intelligible audio signal for the audio dataset used.
Recommended Citation
Faria, Rafael; Earl, Liam; and Gall, Connor, "COMPARATIVE ANALYSIS OF BASIC DENOISING METHODS FOR SPEECH IMPROVEMENT USING STATISTICAL MEASURES" (2026). 2026 Symposium. 35.
https://dc.ewu.edu/srcw_2026/ps_2026/p1_2026/35
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
COMPARATIVE ANALYSIS OF BASIC DENOISING METHODS FOR SPEECH IMPROVEMENT USING STATISTICAL MEASURES
PUB NCR
This research presents a statistical analysis on the effectiveness of four audio denoising techniques: Spectral Gating, Butterworth Filtering, Gaussian Filtering, and Wavelet Denoising. We utilize the signal-to-noise ratio of speech signals (SNR), Root Mean Square Error (RMSE), and Short-Time Objective Intelligibility (STOI) as statistical metrics to compare the effectiveness of the denoising techniques. After statistical comparison we establish that the Spectral Gating denoising technique is superior to the other techniques in denoising and producing an intelligible audio signal for the audio dataset used.
