通过卷积层的特征图来理解对抗性强度.
IEEE transactions on neural networks and learning systems
|February 8, 2024
概括
在平均聚合之前更大的特征地图增强了神经网络的对抗性强度. 修改下方采样操作员可以提高对干扰的抵抗力,而不会牺牲准确性.
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:
- 神经网络中的对抗性强度对于可靠的AI系统至关重要.
- 稳固性取决于模型容量和反扰动能力.
- 卷积层中的特征图是理解反扰动能力的关键.
研究的目的:
- 为了研究卷积特征地图大小和对抗性强度之间的关系.
- 确定聚合操作如何影响这种关系.
- 提出简单的修改来提高神经网络的稳定性.
主要方法:
- 在平均聚合和最大聚合之前,对特征图的理论分析.
- 提出诸如输入上样和下样运营商的步骤调整等修改.
- 在基准架构 (AlexNet,VGG,ResNet18,PreActResNet18) 上进行实证验证.
主要成果:
- 在平均聚合之前更大的特征地图与更好的抗扰性相关.
- 这种相关性不适用于最大池.
- 拟议的修改导致了自然准确性和对抗性强度的显著改善.
结论:
- 在平均聚合之前的特征地图大小是对抗强度的关键因素.
- 简单的架构修改可以提高神经网络的弹性.
- 研究结果为设计更强大的神经网络提供了实用的见解.
相关概念视频
Convolution Properties I
151
Convolution computations can be simplified by utilizing their inherent properties.
The commutative property reveals that the input and the impulse response of an LTI (Linear Time-Invariant) system can be interchanged without affecting the output:
The commutative property reveals that the input and the impulse response of an LTI (Linear Time-Invariant) system can be interchanged without affecting the output:
151
Convolution Properties II
199
The important convolution properties include width, area, differentiation, and integration properties.
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
The width property indicates that if the durations of input signals are T1 and T2, then the width of the output response equals the sum of both durations, irrespective of the shapes of the two functions. For instance, convolving two rectangular pulses with durations of 2 seconds and 1 second results in a function with a width of 3 seconds.
The area property asserts that the area under the...
199
Difference from Background: Limit of Detection
6.4K
The limit of detection (LOD) is the smallest amount of analyte that can be distinguished from the background noise. The LOD value corresponds to the concentration at which the analyte signal is three times larger than the standard deviation of the blank signal. Below this value, the analyte signal cannot be differentiated from the background noise. It is calculated by dividing the calibration slope by 3 times the standard deviation of the blank signals.
The LOD indicates the presence or absence...
The LOD indicates the presence or absence...
6.4K
Deconvolution
160
Deconvolution, also known as inverse filtering, is the process of extracting the impulse response from known input and output signals. This technique is vital in scenarios where the system's characteristics are unknown, and they must be inferred from the observable signals.
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...
160
Residuals and Least-Squares Property
7.4K
The vertical distance between the actual value of y and the estimated value of y. In other words, it measures the vertical distance between the actual data point and the predicted point on the line
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
7.4K
Convolution: Math, Graphics, and Discrete Signals
257
In any LTI (Linear Time-Invariant) system, the convolution of two signals is denoted using a convolution operator, assuming all initial conditions are zero. The convolution integral can be divided into two parts: the zero-input or natural response and the zero-state or forced response, with t0 indicating the initial time.
To simplify the convolution integral, it is assumed that both the input signal and impulse response are zero for negative time values. The graphical convolution process...
To simplify the convolution integral, it is assumed that both the input signal and impulse response are zero for negative time values. The graphical convolution process...
257


