}

Home: Applications > Convolutional Neural Networks (CNN)

Convolutional Neural Networks (CNN)

Key to Machine Learning with Data-Intensive Tasks

Today’s real-time systems have an ever-increasingly challenging task – process all the data and make intelligent decisions about what to do next. Convolutional neural networks (CNN) are key to processing all of this data very quickly, and Tensilica® processors and DSPs are ideal because they can be finely tuned to efficiently execute the most demanding CNN algorithms.

CNNs are used in variety of areas, including image and pattern recognition, speech recognition, natural language processing, and video analysis. From smartphones to smart watches, from advanced driver assistance systems (ADAS) to virtual-reality gaming consoles, drone control, and a host of security devices, the application areas that rely on high-resolution imaging (1080p, 4K, and beyond) are growing.

Today, high-resolution imaging is so sophisticated that we’re relying on it for everything from face detection in smartphones to face recognition in security systems to traffic sign recognition in our vehicles, and for the autonomous vehicle of the future (see Tensilica Vision DSP). There’s a great opportunity to use CNN techniques to further enhance computer vision applications to achieve high level of accuracy.

A neural network is a system of interconnected artificial “neurons” that exchange messages between each other. The connections have numeric weights that are tuned during the training process, so that a properly trained network will respond correctly when presented with an image or pattern to recognize. The network consists of multiple layers of feature-detecting “neurons”. Each layer has many neurons that respond to different combinations of inputs from the previous layers.

Figure 1. An artificial neural network

Convolutional neural networks (CNN) are special cases of the neural network described above. A CNN consists of one or more convolutional layers, often with a subsampling layer, which are followed by one or more fully connected layers as in a standard neural network. 

With the advancement in the semiconductor technology to lower geometries and low power highly parallel processors there is an opportunity to provide convolution neural network (CNN) based computer vision solutions.