The volume of “free” data on the internet has been key to the current success
of deep learning. However, it also raises privacy concerns about the
unauthorized exploitation of personal data for training commercial models. It
is thus crucial to develop methods to prevent unauthorized data exploitation.
This paper raises the question: emph{can data be made unlearnable for deep
learning models?} We present a type of emph{error-minimizing} noise that can
indeed make training examples unlearnable. Error-minimizing noise is
intentionally generated to reduce the error of one or more of the training
example(s) close to zero, which can trick the model into believing there is
“nothing” to learn from these example(s). The noise is restricted to be
imperceptible to human eyes, and thus does not affect normal data utility. We
empirically verify the effectiveness of error-minimizing noise in both
sample-wise and class-wise forms. We also demonstrate its flexibility under
extensive experimental settings and practicability in a case study of face
recognition. Our work establishes an important first step towards making
personal data unexploitable to deep learning models.

Go to Source of this post
Author Of this post: <a href="http://arxiv.org/find/cs/1/au:+Huang_H/0/1/0/all/0/1">Hanxun Huang</a>, <a href="http://arxiv.org/find/cs/1/au:+Ma_X/0/1/0/all/0/1">Xingjun Ma</a>, <a href="http://arxiv.org/find/cs/1/au:+Erfani_S/0/1/0/all/0/1">Sarah Monazam Erfani</a>, <a href="http://arxiv.org/find/cs/1/au:+Bailey_J/0/1/0/all/0/1">James Bailey</a>, <a href="http://arxiv.org/find/cs/1/au:+Wang_Y/0/1/0/all/0/1">Yisen Wang</a>

By admin