An ECG Denoising Method Based on the Generative Adversarial Residual Network
Table 1
Important mathematical symbols and their meanings in this paper.
Mathematical symbol
Description
A signal containing noise
Clean signal
Noise
A signal with noise represented by a one-dimensional vector
A clean signal represented by a one-dimensional vector
Noise represented by a one-dimensional vector
Number of samples
The length of a sample
Distribution of the original clean data
The distribution of the generated data for the generator
Original clean signal
The signal generated by the generator
A signal containing noise
Noise
The signal generated by the generator
The discriminator network determines the probability that the signal comes from the original clean signal rather than the signal generated by the generator
Generate the original loss function of the network
The global difference function represents the global difference between the generated signal and the original signal
The local maximum difference function represents the maximum difference between the generated signal and the original clean signal to represent the local difference
Coefficient of
Coefficient of
The probability that the signal is the original clean signal is obtained from the discriminator network
The loss function of the generator network
The loss function of the discriminator network
The underlying mapping value of the expected output
Residual
The input of the residual block
Convolved with operation, to adjust the dimensions of
