光网络是一种轻便的、通用的和纯基于Matlab的深度学习框架。其思想是为深度学习研究提供一个易于理解、易于使用和高效的计算平台。实现此框架可以支持大部分深度学习架构例如多层感知器网络(MLP)、卷积神经网络(CNN)和循环神经网络(RNN)。该框架支持CPU和GPU无缝转换。实验证明此框架在计算机视觉、自然语言处理以及机器人技术中有不同的应用。
The optical network is a lightweight,versatile and purely Matlabbased deep learning framework.The idea underlying its design is to provide an easy-to-understand,easy-to-use and efficient computational platform for deep learning research.The framework supports major deep learning architectures such as multilayer perceptron(MLP)networks,convolutional neuralnetworks(CNN)and recurrent neural networks(RNN).The framework also supports the straightforward switch between bothCPU and GPU for computation.The experiment result indicates that the framework has different applications in computer vision,natural language processing and robotic technology.