Giters

Hadisalman / smoothing-adversarial

Code for our NeurIPS 2019 *spotlight* "Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers"

Home Page:https://arxiv.org/abs/1906.04584

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Do we need to clip our new input into [0,1] after adding noise ?

Alan-Qin opened this issue · comments

commented

Hi @Alan-Qin!

No, no clipping happens after adding noise. Mathematically, the smoothing operation doesn't involve any clipping, and there is no reason why practically this should be done. Hope this answers your question!

commented

Thanks!