inarizuuuushi https://blog.hatena.ne.jp/inarizuuuushi/ Sabrou-mal サブロウ丸 https://inarizuuuushi.hatenablog.com/ Adversarial example ... 入力にごくわずかな摂動(ノイズ)を加えてモデルの出力を狂わせたもの. 特にneural networkモデルに対してadversarial exampleが存在しやすい. このAdversarial exampleを"検出する"という観点の研究をまとめました. Detecting adversarial example from Nariaki Tateiwa 参考 [1] Xu, Weilin, David Evans, and Yanjun Qi. "Feature squeezing: Detecting adversarial examp… 190 <iframe src="https://hatenablog-parts.com/embed?url=https%3A%2F%2Finarizuuuushi.hatenablog.com%2Fentry%2F2021%2F07%2F30%2F204237" title="Detecting Adversarial Examples - Sabrou-mal サブロウ丸" class="embed-card embed-blogcard" scrolling="no" frameborder="0" style="display: block; width: 100%; height: 190px; max-width: 500px; margin: 10px 0px;"></iframe> https://cdn-ak.f.st-hatena.com/images/fotolife/i/inarizuuuushi/20210730/20210730005023.png Hatena Blog https://hatena.blog 2021-07-30 20:42:37 rich https://inarizuuuushi.hatenablog.com/entry/2021/07/30/204237 1.0 100%