Efficient Image Feature Extraction using Convolutional Neural Networks
Aishwary Awasthi, Srikrishna Baskar Rao, Kalyan Acharjya · 2024
Convolutional Neural Networks (CNNs) are a kind of neural network that has grown to be increasingly famous for image-associated responsibilities consisting of object popularity, image segmentation, and picture type. They may be widely utilized in computer vision packages, including characteristic extraction, where the entered photo is scanned to identify and extract functions. CNNs use convolutional layers, which permit the network to discover a specific sample within the input picture. The layers have a fixed of weights that are adjusted primarily based on the input statistics. Those weights are then used to become aware of the functions in the entered picture. CNNs paintings by way of taking in a photograph as enter and multiplying it detail-sensible with a fixed of filters, or kernels. Each of those kernels has a weight that adjusts primarily based on the activation of neurons in the network. The output is a function map that indicates the presence of a specific sample in the picture. This option map is then used to discover the features inside the photo.