Structure verification of deep neural networks at compilation time

Abstract

This paper presents TensorSafe , a Haskell library that makes possible the definition and structural validation of deep neural network architectures. Nowadays, the development process of deep learning models has been notably simplified due to the availability of sophisticated tools in the industry. However, most of these tools do not provide any security controls at compilation time, making the developers deal with unexpected run-time errors and uncertainties. In particular, validating the structure of deep neural networks at compilation time is a complex subject that involves the mathematical validation of all operations that a deep learning model will perform. Moreover, this structural checking requires an advanced usage of types systems theories to manipulate abstract type definitions capable of modeling neural network constructions. Many different programming techniques were involved in the specification of TensorSafe. Primarily, the application of the functional programming paradigm and the use of type-level programming were of great importance for the development process and to probe the correctness of the neural network models. The experimental evaluation showed that by using TensorSafe it is possible to correctly create well-known deep neural network architectures, such as AlexNet or ResNet50 . • Review state of the art of deep learning DSLs that provide structural controls at compile time. • Extensible Haskell library that is capable of defining non-sequential deep learning architectures. • Interaction with external deep learning frameworks like TensorFlow for JavaScript or Keras for Python.