MatConvNet: CNNs for MATLAB

MatConvNet is a MATLAB toolbox implementing Convolutional Neural Networks (CNNs) for computer vision applications. It is simple, efficient, and can run and learn state-of-the-art CNNs. Many pre-trained CNNs for image classification, segmentation, face recognition, and text detection are available.

New: 1.0-beta24 released with bugfixes, new examples, and utility functions.

New: 1.0-beta23 released with vl_nnroipool and a Fast-RCNN demo.

New: 1.0-beta22 released with a few bugfixes.

New: 1.0-beta21 provides two new tools, vl_tmove and ParameterServer, to accelerate significantly data transfers between multiple GPUs. It also provides a new version of vl_imreadjpeg that allows to load, transform, and transfer data to the GPU in parallel, resulting in significant speedups in training and testing (20% to 400%, depending on the model). vl_nnconv now has a dilate option for dilated convolution.

New: 1.0-beta20 adds the binlinear resampler layer vl_nnbilinearsampler and a spatial transformer example.

New: 1.0-beta19 adds pre-trained ResNet models (demo training code coming next), CuDNN V5 support, and numerous other improvements and bugfixes.

Obtaining MatConvNet


Getting started

Use cases

Other information