CNN Convolutional Layer Operations 3 — Questions and Answers
Question 1: What is the receptive field of a neuron in the output of a convolutional layer?
- The region of the input that influences that neuron's activation (Correct answer)
- The size of the filter
- The number of input channels
- The output feature map dimensions
Correct answer: The region of the input that influences that neuron's activation
The receptive field is the spatial region in the input (or previous feature map) that a given output neuron 'sees' and is influenced by.
Question 2: How does increasing filter size (e.g., from 3×3 to 7×7) affect the receptive field of a single convolutional layer?
- Larger receptive field per layer (Correct answer)
- Smaller receptive field per layer
- No change to receptive field
- Receptive field becomes equal to input size
Correct answer: Larger receptive field per layer
A larger filter directly covers more spatial area of the input, increasing the receptive field of each output neuron in that layer.
Question 3: In a depthwise separable convolution, what does the depthwise step do?
- Applies a separate filter to each input channel independently (Correct answer)
- Combines all channels into one output
- Applies one filter across all channels
- Performs 1×1 convolutions only
Correct answer: Applies a separate filter to each input channel independently
The depthwise step convolves each input channel with its own dedicated spatial filter, keeping channels separate.
Question 4: What is the output depth (number of channels) when 128 filters are applied to a 64-channel input?
- 128 (Correct answer)
- 64
- 8192
- 192
Correct answer: 128
The output depth always equals the number of filters, regardless of the input channel count.
Question 5: Which padding strategy discards border activations and produces a smaller output than the input (with stride=1)?
- Valid padding (Correct answer)
- Same padding
- Full padding
- Causal padding
Correct answer: Valid padding
Valid (no) padding does not add zeros, so the filter cannot center on border pixels, shrinking the output.
Question 6: A 1×1 convolution applied to a feature map with C channels produces K output channels. What is its primary use?
- Channel-wise linear combination to change depth without altering spatial size (Correct answer)
- Spatial downsampling
- Edge detection
- Batch normalization replacement
Correct answer: Channel-wise linear combination to change depth without altering spatial size
1×1 convolutions mix channel information linearly, effectively projecting to a different channel dimension while preserving spatial resolution.
Question 7: During the forward pass, what is the bias term in a convolutional layer added to?
- Each output feature map value (one bias per filter) (Correct answer)
- Each input pixel
- The filter weights before convolution
- The loss function directly
Correct answer: Each output feature map value (one bias per filter)
One scalar bias per filter is added to every spatial position in that filter's output feature map.
CNN Convolutional Layer Operations 3
What is the receptive field of a neuron in the output of a convolutional layer?