A convolutional neural network is a neural-network architecture that uses convolutional filters to detect local patterns in data such as images, video frames, audio, or signals.
CNNs are strongly associated with computer vision because they can learn edges, textures, shapes, and higher-level visual features through stacked layers.