{"description":"Evaluate your understanding of pooling operations, feature map visualization, and the role of kernel design in CNNs.","questions":[{"answer":"To reduce spatial dimensions and computational load","number":1,"options":["To increase the number of filters","To apply activation functions","To reduce spatial dimensions and computational load","To normalize input images"],"question":"What is the primary purpose of pooling layers in a CNN?"},{"answer":"Max pooling","number":2,"options":["Average pooling","Max pooling","Min pooling","Sum pooling"],"question":"Which pooling technique selects the highest value in a region of the feature map?"},{"answer":"2x2","number":3,"options":["1x1","2x2","4x4","7x7"],"question":"What is a typical size for a pooling filter in CNNs?"},{"answer":"It reduces overfitting by keeping only dominant features","number":4,"options":["It reduces overfitting by keeping only dominant features","It increases training time significantly","It lowers model accuracy","It removes convolution layers"],"question":"How does max pooling affect a CNN\u2019s ability to generalize?"},{"answer":"To understand what the network is learning at each layer","number":5,"options":["To visualize class probabilities","To improve loss functions","To understand what the network is learning at each layer","To speed up pooling operations"],"question":"Why is visualizing feature maps useful in CNNs?"},{"answer":"Edges and textures","number":6,"options":["Entire objects","Object categories","Edges and textures","Fully connected outputs"],"question":"What type of features are typically captured in the early convolutional layers?"},{"answer":"Faster computation and fewer parameters","number":7,"options":["Increased image resolution","Faster computation and fewer parameters","Higher memory usage","Improved color reproduction"],"question":"Which of the following is a benefit of reducing the size of feature maps?"},{"answer":"Filter size","number":8,"options":["Learning rate","Activation function","Filter size","Batch size"],"question":"In CNNs, what determines the area of the image a pooling filter covers?"},{"answer":"Vertical edge detector","number":9,"options":["Horizontal edge detector","Emboss filter","Vertical edge detector","Identity filter"],"question":"Which kernel would be most useful for detecting vertical lines in an image?"},{"answer":"Finding the right pattern and size to detect desired features","number":10,"options":["Ensuring the kernel is stored as a JPEG","Deciding the number of epochs","Finding the right pattern and size to detect desired features","Choosing the correct font for visualization"],"question":"What is a common challenge when designing custom convolution kernels?"}],"title":"Pooling Layers in CNNs"}