240 发简信
IP属地:浙江
  • JPEG-resistant Adversarial Images

    摘要: While JPEG compression is not differentiable, we show how to closely...

  • 研读笔记-COUNTERING ADVERSARIAL IMAGES USING INPUT TRANSFORMATIONS

    文章链接:http://arxiv.org/pdf/1711.00117项目源码:https://github.com/facebookrese...

  • Resize,w 360,h 240
    研读笔记-A study of the effect of JPG compression on adversarial images

    文章链接:http://arxiv.org/pdf/1608.00853 摘要:文中评估JPG压缩对对抗性图像分类的影响。对于小幅度的FGSM扰...

  • Resize,w 360,h 240
    研读笔记-Universal adversarial perturbations

    文章链接:https://arxiv.org/abs/1610.08401 摘要:提出一个系统的方法可以生成 universal perturb...

  • 研读笔记-improving the robustness of deep neural networks via stability training

    Related work Data augmentation:提高模型识别能力的方法之一是增大样本数据,往往更大的样本数据集可以训练出表现更好的...