Publications by authors named "Qingbin Du"
Sensors (Basel)
November 2024
Article Synopsis
- This study introduces a deep-learning denoising method that transforms fiber-optic sensor spectra into 2D images and uses a Cycle-GAN model to improve signal quality.
- It demonstrates significant improvements in signal-to-noise ratio (SNR), root mean square error (RMSE), and high correlation with original signals compared to traditional denoising methods like wavelet transform and empirical mode decomposition.
- The proposed algorithm successfully reduces noise across different fiber-optic sensors and shows excellent linearity in temperature response, making fiber-optic sensing more effective for various research and industrial applications.
View Article and Find Full Text PDF