Denoising Autoencoders
Reconstruct clean data from corrupted input.
From Copying to Cleaning
A plain autoencoder rebuilds its input. A denoising autoencoder goes further: you feed it a corrupted version and ask it to restore the clean original.
Add Noise on Purpose
You deliberately damage the input with noise before feeding it in. The clean original stays as the target the network must reach.
noisy = clean + 0.3 * torch.randn_like(clean)