Convolutional neural networks (CNNs, or ConvNets) are essential tools for deep learning, and are especially useful for image classification, object detection, and recognition tasks. CNNs are implemented as a series of interconnected layers. The layers are made up of repeated blocks of convolutional, ReLU (rectified linear units), and pooling layers. The convolutional layers convolve their input with a set of filters. The filters were automatically learned during network training. The ReLU layer adds nonlinearity to the network, which enables the network to approximate the nonlinear mapping between image pixels and the semantic content of an image. The pooling layers downsample their inputs and help consolidate local image features.
Convolutional neural networks require Deep Learning Toolbox™. Training and prediction are supported on a CUDA®-capable GPU with a compute capability of 3.0 or higher. Use of a GPU is recommended and requires Parallel Computing Toolbox™.
You can construct a CNN architecture, train a network using semantic segmentation, and use the trained network to predict class labels or detect objects. You can also extract features from a pretrained network, and use these features to train a classifier. Additionally, you can perform transfer learning which retrains the CNN on new data.You can also use the Image Labeler, Video Labeler, feature extractors, and the Deep Learning Toolbox classifiers to create a custom detector.
|Train an R-CNN deep learning object detector|
|Train a Fast R-CNN deep learning object detector|
|Train a Faster R-CNN deep learning object detector|
|Detect objects using R-CNN deep learning detector|
|Detect objects using Fast R-CNN deep learning detector|
|Detect objects using Faster R-CNN deep learning detector|
|ROI input layer for Fast R-CNN|
|Neural network layer used to output fixed-size feature maps for rectangular ROIs|
|Softmax layer for region proposal network (RPN)|
|Classification layer for region proposal networks (RPNs)|
|Box regression layer for Fast and Faster R-CNN|
|Region proposal layer for Faster R-CNN|
R-CNN, Fast R-CNN, and Faster R-CNN basics
Deep Learning in MATLAB (Deep Learning Toolbox)
Discover deep learning capabilities in MATLAB® using convolutional neural networks for classification and regression, including pretrained networks and transfer learning, and training on GPUs, CPUs, clusters, and clouds.