相关实验视频
尼斯特罗夫加速梯度的优势在于提高了隐蔽的敌对攻击的可转移性
1Joint Laboratory of Data Science and Business Intelligence, Southwestern University of Finance and Economics, Chengdu, Sichuan, China.
PloS one
|November 25, 2025
概括
本研究介绍了Diff-AdaNAG,这是一种用于深度神经网络创建隐形对抗示例的新方法. 它提高了黑子攻击中的可转移性,而不会牺牲不可察觉的干扰.
科学领域:
- 人工智能的人工智能
- 机器学习 机器学习
- 计算机视觉 计算机视觉
背景情况:
- 深度神经网络容易受到对抗性示例的影响,这些示例是带有微妙扰动的输入,旨在误导模型.
- 对抗性攻击的关键指标是可转移性 (跨不同模型的成功) 和隐蔽性 (不可察觉的干扰).
- 在可转移性和隐形性之间存在冲突;高的可转移性往往意味着明显的噪音,而隐形的例子在黑盒设置中表现不佳.
研究的目的:
- 提出一个新的框架,Diff-AdaNAG,它解决了在对抗性示例生成中可转移性和隐蔽性之间的冲突.
- 为了提高在黑盒场景中的对抗性攻击的性能,同时保持不可察觉的干扰.
主要方法:
- Diff-AdaNAG框架将Nesterov的加速梯度 (NAG) 集成到基于扩散的对抗示例生成中.
- 一个扩散机制将示例生成引导到自然数据分布,确保隐形性.
- 一个自适应的步骤大小策略利用了NAG的加速和通用化,以改善黑子可转移性.
主要成果:
- 在白盒和黑盒对抗性攻击场景中,Diff-AdaNAG始终超过现有的最先进的方法.
- 该框架显著提高了对抗性示例的可转移性.
- 保持了隐蔽性,敌对的例子仍然是不可察觉的.
结论:
- 在对抗性攻击中,diff-AdaNAG有效地解决了隐形性和可转移性之间的权衡.
- 拟议的方法在为深度神经网络生成有效和隐蔽的对抗示例方面取得了重大进展.
- 该框架在受控 (白盒) 和现实 (黑盒) 攻击环境中都表现出卓越的性能.
相关概念视频
02:07Improving Translational Accuracy
14.0K
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
14.0K
02:07Improving Translational Accuracy
3.5K
3.5K
01:30Acceleration Vectors
21.6K
In everyday conversation, accelerating means speeding up. Acceleration is a vector in the same direction as the change in velocity, Δv, therefore the greater the acceleration, the greater the change in velocity over a given time. Since velocity is a vector, it can change in magnitude, direction, or both. Thus acceleration is a change in speed or direction, or both. For example, if a runner traveling at 10 km/h due east slows to a stop, reverses direction, and continues their run at 10 km/h...
21.6K
01:24Confidence Coefficient
10.3K
The confidence coefficient is also known as the confidence level or degree of confidence. It is the percent expression for the probability, 1-α, that the confidence interval contains the true population parameter assuming that the confidence interval is obtained after sufficient unbiased sampling; for example, if the CL = 90%, then in 90 out of 100 samples the interval estimate will enclose the true population parameter. Here α is the area under the curve, distributed equally under...
10.3K
02:18Types of Genetic Transfer Between Organisms
30.5K
Genetic transfer occurs when genetic information is passed from one organism to another. It occurs via two mechanisms: vertical gene transfer and horizontal gene transfer. Vertical gene transfer occurs when genetic information is transferred from one generation to the next, which happens much more frequently than horizontal gene transfer. Both sexual and asexual reproduction are forms of vertical gene transfer, where one or more organisms pass some or all of their genome onto their progeny.
30.5K
02:18Types of Genetic Transfer Between Organisms
6.1K
6.1K