Introduction: We introduce a deep learning framework, named FudanDNN, developed by Shanghai Key Laboratory of Intelligent Information Processing, at Fudan University. Unlike other tools, FudanDNN is designed to be used in business environments, rather than as a research tool. It is a C++-based, industry-focused, distributed deep-learning framework made with expression, speed, and modularity. FudanDNN neural networks include Restricted Boltzmann machines (RBM), autoencoders, Convolutional Neural Networks (CNN), Recursive Neural Networks (LSTMs) and any combination of them, which allows for fast prototyping for non-experts. The most distinctive feature of this tool is its GUI-based designer. By this designer, the structure of deep networks can be easily defined by simple drag-and-drop operations.
Team Member: Xiaoqing Zheng, Siyan Li, Jiangtao Feng, Mengxiao Lin, Hao Song, Shangtong Zhang
Affiliation: Shanghai key laboratory of intelligent information processing, Fudan University
Email: zhengxq@fudan.edu.cn