MATConvNet is a MATLAB toolbox for implementing Convolutional Neural Networks (CNNs). It provides a simple and flexible interface for building and training CNNs, with functions for linear convolutions, feature pooling, and other essential operations. The toolbox supports efficient computation on both CPU and GPU, enabling the training of complex models on large datasets like ImageNet. It includes a comprehensive documentation, examples, and technical details of each computational block. The toolbox is designed for ease of use, with a focus on simplicity and efficiency, and allows researchers to quickly prototype and test new CNN architectures. MATConvNet includes various computational blocks such as convolution, deconvolution, pooling, activation functions, normalization, and categorical losses. It also provides pre-trained models for image classification, segmentation, text spotting, and face recognition. The toolbox supports both simple and complex network structures, including sequences and directed acyclic graphs (DAGs), and includes functions for backpropagation to compute derivatives and optimize CNNs. MATConvNet is open-source and can be used for research and development in computer vision and deep learning.MATConvNet is a MATLAB toolbox for implementing Convolutional Neural Networks (CNNs). It provides a simple and flexible interface for building and training CNNs, with functions for linear convolutions, feature pooling, and other essential operations. The toolbox supports efficient computation on both CPU and GPU, enabling the training of complex models on large datasets like ImageNet. It includes a comprehensive documentation, examples, and technical details of each computational block. The toolbox is designed for ease of use, with a focus on simplicity and efficiency, and allows researchers to quickly prototype and test new CNN architectures. MATConvNet includes various computational blocks such as convolution, deconvolution, pooling, activation functions, normalization, and categorical losses. It also provides pre-trained models for image classification, segmentation, text spotting, and face recognition. The toolbox supports both simple and complex network structures, including sequences and directed acyclic graphs (DAGs), and includes functions for backpropagation to compute derivatives and optimize CNNs. MATConvNet is open-source and can be used for research and development in computer vision and deep learning.