在 2D 卷积神经网络（CNN）中，卷积核的数量说明了什么？

In a 2-D convolutional neural network (CNN), what does the number of kernels define?

Question

在 2D 卷积神经网络（CNN）中，卷积核的数量说明了什么？

In a 2-D convolutional neural network (CNN), what does the number of kernels define?

BlackTom AI · Accepted Answer

Each convolutional kernel produces one output channel (feature map); therefore the number of kernels equals the number of output feature maps.

1) Relevant knowledge points and concepts
- A 2-D convolutional layer applies multiple kernels (filters) over the input volume.  
- Each kernel has shape (kH, kW, C_in) where C_in is the number of input channels.  
- Sliding a kernel over spatial positions computes a dot product at each location, producing a 2-D map. That map is one output channel (feature map).  
- The layer’s output shape is (N, H_out, W_out) where N is the number of kernels (sometimes called out_channels).

2) Step-by-step reasoning
- You choose N kernels for the layer.  
- For each kernel i (i = 1..N): convolve it with the full input (all input channels) → produce a 2-D activation map (feature map) for kernel i.  
- Stack these N maps along the channel dimension → produce N feature maps (the layer’s output channels).

3) Calculations and intermediate steps (example)
- Example: input 32×32 RGB image: H=32, W=32, C_in=3. Kernel size K=3, number of kernels N=64, stride=1, no padding (valid).  
- Each kernel shape: 3×3×3 => 27 weights + bias.  
- For one spatial position: one dot product = 27 multiplications + sums → single scalar.  
- H_out = H − K + 1 = 30, W_out = W − K + 1 = 30.  
- One kernel produces 30×30 = 900 scalars → one 30×30 feature map.  
- All kernels produce N × H_out × W_out = 64 × 30 × 30 values, i.e., 64 feature maps.

4) Why the correct answer is correct
- By definition each learned kernel extracts one map of activations across spatial locations; thus the number of kernels directly sets how many distinct feature maps (output channels) the layer produces.

5) Why the other options are incorrect
- “输入像素的数量 / The number of input pixels”: kernel count does not change the number of input pixels.  
- “层数 / The number of layers”: kernel count is per layer and does not indicate how many layers exist.  
- “输出值的数量 / The number of output values”: kernel count alone is not the total number of output values (that depends also on H_out and W_out). Kernel count equals the number of output channels (feature maps), not the full scalar count.

6) Conclusion
The number of kernels defines the number of feature maps (output channels). This matches the provided correct answer: 特征图的数量 The number of feature maps.

在 2D 卷积神经网络（CNN）中，卷积核的数量说明了什么？ In a 2-D convolutional neural network (CNN), what does the number of kernels define?Single choice

Similar Questions

How can deep convolutional neural network (DCNN) face recognition models be more like human face processing?

Which statement best describes what early CNN layers tend to learn?

In a deep neural network (e.g., a CNN for image classification), hidden layers often learn a feature hierarchy. Which statement best describes what this means?

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

More Practical Tools for Students Powered by AI Study Helper

Homework AI Solver

Stylized AI Paper Writer

Plagiarism Checker Assistant

Citation AI Academic Writing Tool

In-Class Translation Assistant

AI Note Generator

AI Quiz Answers

Past Exam Questions from University Test Bank

Smart Practice Assistant

Adaptive Practice

Making Your Study Simpler