Fetching the paper…

Invisible Backdoor Attacks on Deep Neural Networks via Steganography and Regularization · Around