相关实验视频
Updated: Jul 2, 2025

04:43
Visualizing Visual Adaptation
Published on: April 24, 2017
9.0K
基于颜色空间转换的对抗性攻击的防御
Haoyu Wang1, Chunhua Wu1, Kangfeng Zheng1
1School of Cyberspace Security, Beijing University of Posts and Telecommunications, China.
概括
这项研究引入了一种针对深度学习模型的对抗性攻击的新防御方法. 通过放大对抗性扰动,该方法使攻击变得可见,并使其无效.
科学领域:
- 计算机科学 计算机科学
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 深度学习 (DL) 模型实现了高性能,但容易受到对抗性攻击.
- 敌对的例子,微妙的输入干扰,可能导致错误分类.
- 现有的研究表明,完全消除对抗性例子是不可能的.
研究的目的:
- 通过取消现有的例子和扩大新的例子来开发对抗对手攻击的防御方法.
- 增强深度学习模型对抗对抗干扰的稳定性.
- 为了使人类观察者能够区分对抗性干扰.
主要方法:
- 通过改变分类界限,使现有的对抗性示例无效.
- 增加对新生成示例的对抗性干扰,以提高可见性.
- 通过颜色空间转换实现防御策略.
主要成果:
- 在CIFAR-10,CIFAR-100和Mini-ImageNet数据集上证明了拟议的防御方法的有效性和多功能性.
- 成功地使现有的对抗性示例无效.
- 让对抗性扰动更容易被人类感知.
结论:
- 基于放大对抗性扰动的提议防御方法,对抗性攻击有效.
- 这代表了深度学习中对抗防御的新方法.
- 该方法为提高人工智能系统的稳定性提供了一个实际的解决方案.
相关概念视频
Color Vision
578
Color perception begins in the retina, the light-sensitive layer at the back of the eye. Two main theories explain how colors are seen: the trichromatic theory and the opponent-process theory. The trichromatic theory, proposed by Thomas Young in 1802 and extended by Hermann von Helmholtz in 1852, suggests that color vision is based on three types of cone receptors in the retina. These cones are sensitive to different but overlapping ranges of wavelengths corresponding to red, blue, and green.
578
Photoreceptors and Visual Pathways
6.0K
At the molecular level, visual signals trigger transformations in photopigment molecules, resulting in changes in the photoreceptor cell's membrane potential. The photon's energy level is denoted by its wavelength, with each specific wavelength of visible light associated with a distinct color. The spectral range of visible light, classified as electromagnetic radiation, spans from 380 to 720 nm. Electromagnetic radiation wavelengths exceeding 720 nm fall under the infrared category,...
6.0K

