A type of deep neural network optimized for analyzing visual imagery, consisting of various layers including convolutional, pooling, and fully connected layers. These networks employ convolution operations to detect hierarchical patterns and pooling to reduce spatial dimensions, thereby enhancing computational efficiency. CNNs are extensively used in image and video recognition, image classification, and tasks requiring visual data analysis.