隐形车辆对抗性伪装纹理生成基于神经风格转移的神经风格转移
Wei Cai1, Xingyu Di1, Xin Wang1
1The Third Faculty of Xi'an Research Institute of High Technology, Xi'an 710064, China.
Entropy (Basel, Switzerland)
|November 27, 2024
概括
研究人员开发了对抗性伪装纹理,可以在物理世界中欺骗深度神经网络 (DNN). 这些纹理是隐蔽的,在特定场景中逃避DNN和人类观察者.
科学领域:
- 计算机视觉 计算机视觉
- 人工智能的人工智能
- 深度学习 (Deep Learning) 是一种深度学习.
背景情况:
- 深度神经网络 (DNN) 在数字和物理环境中都容易受到敌对攻击.
- 现有的对抗性伪装纹理优先考虑攻击的有效性而不是隐形性,对人类观察者来说看起来不自然.
- 这种缺乏隐形性限制了对抗性攻击在现实场景中的实际应用.
研究的目的:
- 开发对抗性伪装纹理,这些纹理对DNN对象探测器有效,并且对人类观察者来说不显眼.
- 通过将风格传输模块集成到生成框架中,提高对抗性攻击的隐形性.
- 创建在数字和物理领域都强大的对抗性纹理.
主要方法:
- 提出了一个新的框架,包括用于对抗性纹理生成的风格传输模块.
- 利用风格损失计算来指导纹理生成,确保与周围环境的视觉连贯性.
- 在数字模拟和物理世界实验中评估生成的对抗性伪装纹理.
主要成果:
- 提出的方法成功地产生了具有高度隐形性的对抗性伪装纹理.
- 这些纹理在数字和物理环境中有效地欺骗了先进的DNN对象探测器.
- 在特定的,有针对性的场景中,对抗性的纹理对人类观察者来说是不显眼的.
结论:
- 整合风格转移显著提高了对手伪装纹理的隐形性.
- 开发的纹理为逃避基于DNN的对象检测提供了一个有希望的解决方案,同时保持视觉微妙性.
- 这种方法促进了对抗性攻击在现实世界计算机视觉系统中的实际应用.
相关概念视频
Aliasing
119
Accurate signal sampling and reconstruction are crucial in various signal-processing applications. A time-domain signal's spectrum can be revealed using its Fourier transform. When this signal is sampled at a specific frequency, it results in multiple scaled replicas of the original spectrum in the frequency domain. The spacing of these replicas is determined by the sampling frequency.
If the sampling frequency is below the Nyquist rate, these replicas overlap, preventing the original...
If the sampling frequency is below the Nyquist rate, these replicas overlap, preventing the original...
119
Transformers
1.1K
A device that transforms voltages from one value to another using induction is called a transformer. A transformer consists of two separate coils, or windings, wrapped around the same soft iron core. However, they are electrically insulated from each other.
The iron core has a substantial relative permeability. Therefore, the magnetic field lines generated due to the current in one winding are almost entirely confined within the core, such that the same magnetic flux permeates each turn of both...
The iron core has a substantial relative permeability. Therefore, the magnetic field lines generated due to the current in one winding are almost entirely confined within the core, such that the same magnetic flux permeates each turn of both...
1.1K
Nonconscious Mimicry
4.5K
Nonconscious mimicry occurs when individuals alter their mannerisms to match the behaviors and expressions of those nearby, without intention.
4.5K
Forced Transdifferentiation
1.9K
Transdifferentiation, also known as lineage reprogramming, was first discovered by Selman and Kafatos in 1974 in silkmoths. They observed that the moths’ cuticle-producing cells transformed into salt-producing cells. Many such cases of natural transdifferentiation occur in organisms. In humans, pancreatic alpha cells can become beta cells. In newts, the loss of the eye’s lens causes the pigmented epithelial cells to transdifferentiate into the lens cells.
Artificial...
Artificial...
1.9K
Motor and Sensory Areas of the Cortex
2.9K
The cerebral cortex, the brain's outermost layer, is pivotal in processing complex cognitive tasks, emotions, and various sensory inputs and executing voluntary motor activities. This intricate structure is divided into three primary functional areas: the motor areas, sensory areas, and association areas.
Motor Areas
The motor areas located in the frontal lobe are central to controlling voluntary movements. This region is further subdivided into the primary motor cortex and the premotor...
Motor Areas
The motor areas located in the frontal lobe are central to controlling voluntary movements. This region is further subdivided into the primary motor cortex and the premotor...
2.9K
Gradient and Del Operator
2.5K
In mathematics and physics, the gradient and del operator are fundamental concepts used to describe the behavior of functions and fields in space. The gradient is a mathematical operator that gives both the magnitude and direction of the maximum spatial rate of change. Consider a person standing on a mountain. The slope of the mountain at any given point is not defined unless it is quantified in a particular direction. For this reason, a "directional derivative" is defined, which is a vector...
2.5K


