MatConvNet: Convolutional Neural Networks for MATLAB

By Andrea Vedaldi, Karel Lenc. 2014

2,983 citations recorded by the source corpus.

MatConvNet is an open source implementation of Convolutional Neural Networks (CNNs) with a deep integration in the MATLAB environment. The toolbox is designed with an emphasis on simplicity and flexibility. It exposes the building blocks of CNNs as easy-to-use MATLAB functions, providing routines for computing convolutions with filter banks, feature pooling, normalisation, and much more. MatConvNet can be easily extended, often using only MATLAB code, allowing fast prototyping of new CNN architectures. At the same time, it supports efficient computation on CPU and GPU, allowing to train complex models on large datasets such as ImageNet ILSVRC containing millions of training examples

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

A Theory of Generative ConvNet (2016, methodology): “The code in our work is based on the Matlab code of MatConvNet of [32].”
Learning feed-forward one-shot learners (2016, positive): “All experiments were performed using MatConvNet [22].”
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation (2015, positive): “Our system is built on MatConvNet [46].”
Learning without Forgetting (2016, methodology): “We use MatConvNet [31] to train our networks using stochastic gradient descent with momentum of 0.”
Attribute2Image: Conditional Image Generation from Visual Attributes (2015, background): “…assumed to follow isotropic multivariate Gaussian distribution, while two conditional distributions pθ(x|y, z) and qφ(z|x, y) are multivariate Gaussian distributions whose mean and covariance are parametrized by N ( µθ(z, y), diag(σ 2 θ(z, y)) ) and N ( µφ(x, y), diag(σ 2 φ(x, y)) ) , respectively.”
UvA-DARE (Digital Academic Repository) Self-Supervised Video Representation Learning With Odd-One-Out Networks (positive): “The self-supervised network is trained with stochastic gradient descent using MatConvNet [40].”
Spatiotemporal Residual Networks for Video Action Recognition (2016, methodology): “4 Model training and evaluation Our method has been implemented in MatConvNet [28] and we share our code and models at https://github.”
Accessorize to a Crime: Real and Stealthy Attacks on State-of-the-Art Face Recognition (2016, methodology): “We used MatConvNet, an NN toolbox for MATLAB, to train DNN B and DNN C , run the DNNs, and test our attacks [41].”
Financial Time Series Prediction Using Deep Learning (2017, methodology): “2 that was trained using the MatConvNet package [32] with mini-batches of size 100, where the learning rate was adaptively reduced, by a factor of 5, when observing a ﬂat error reduction of the validation set during training.”
End-to-End Incremental Learning (2018, neutral): “Our models are implemented on MatConvNet [35].”
VoxCeleb2: Deep Speaker Recognition (2018, methodology): “Our implementation is based on the deep learning toolbox MatConvNet [44].”
VoxCeleb: A Large-Scale Speaker Identification Dataset (2017, methodology): “Our implementation is based on the deep learning toolbox MatConvNet [39] and trained on a NVIDIA TITAN X GPU.”
Scene Flow to Action Map: A New Representation for RGB-D Based Action Recognition with Convolutional Neural Networks (2017, methodology): “For SFAM-CTKRP, we revised the codes of paper [3] based on MatConvNet [54].”
Fast ConvNets Using Group-Wise Brain Damage (2015, methodology): “…majority of current implementations of generalized convolutions (including the most efficient one at the time of submission) [7, 13, 20, 9, 35, 30] compute generalized convolutions by reducing them to matrix multiplications (this reduction is also referred to as lowering, unrolling, or the…”