The recognition of multi-components signals based on semantic segmentation

MIDW‐Net: A multi‐tasking network architecture for radar intra‐pulse parameter description

“…Convolutional denoising autoencoder: Traditional filtering algorithms, such as 2D Wiener filtering [4], the non-local mean image denoising method [5], and the image sparse regularization denoising algorithm based on structural similarity [6], are utilized in electromagnetic environments with low signal-to-noise ratios (SNRs). However, these filtering algorithms can only filter part of the noise [7]. To eliminate more noise and retain signal information, we use a convolutional denoising autoencoder (CDAE) to filter the time-frequency image (TFI) of the radar signal and restore the energy distribution of the original TFI from the image with a high noise level.…”