Convolutional layers are

Question

BlackTom AI · Accepted Answer

The question asks about what convolutional layers are, given multiple possible answers.
Option 1: 'neither' — This is the correct choice in the context of the provided options. Standard convolutional layers are not invariant or equivariant to rotations by default.
Option 2: 'invariant to all rotations' — This is incorrect. A convolution operation does not produce rotation invariance on its own; rotating the input typically yields a different feature map unless additional mechanisms (like data augmentation or specialized rotation-equivariant architectures) are used.
Option 3: 'equivariant to all rotations' — This is also incorrect for vanilla convolutions. A standard convolution is not rotation-equivariant; rotating the input does not simply rotate the output feature map. Equivariance to transformations other than translations requires architectural tweaks or group convolutions, not plain conv layers.

To connect this to a core point: while convolution is famously translation-equivariant (shifting the input results in a corresponding shift in the output), rotation-based properties are not guaranteed by standard convolutions. Therefore, among the given options, 'neither' accurately reflects the default behavior of typical convolutional layers.

CS-7643-O01, OAN, OSZ Quiz #3: Convolutional Neural Network Architectures (Lesson 6), Visualization (Lesson 7), Advanced Computer Vision Architectures (Lesson 8)

Convolutional layers are

查看解析

登录即可查看完整答案

类似问题

Which of the following Neural Networks is used in Machine Vision system?

哪一项陈述正确描述了卷积神经网络 (CNN) 在识别中的特征？ Which statement describes a characteristic of the Convolutional Neural Network (CNN) in recognition correctly?

This is the CNN Architecture diagram that was explained in lectures. Match A, B, C, D, and E to the correct labels below. Note: Full marks will only be awarded for getting all matches correct here. You must make selections in all drop down lists in order for this question to be counted.

In implementing CNNs, we actually implement the forwards pass as a cross-correlation. This means that for the backward pass, the gradient of the loss with respect to the input is

Max pooling layers are:

What is the most typical use for a convolutional neural network?

更多留学生实用工具

考试浏览器助手

风格化写作助手

论文查重助手

文献引用助手

课堂转译助手

课堂笔记助手

Quiz搜索助手

学校历年真题

智能刷题助手

智能匹配练习

希望你的学习变得更简单