Toward Scalable Verification for Safety-Critical Deep Networks

Abstract

The increasing use of deep neural networks for safety-critical applications, such as autonomous driving and flight control, raises concerns about their safety and reliability. Formal verification can address these concerns by guaranteeing that a deep learning system operates as intended, but the state of the art is limited to small systems. In this work-in-progress report we give an overview of our work on mitigating this difficulty, by pursuing two complementary directions: devising scalable verification techniques, and identifying design choices that result in deep learning systems that are more amenable to verification.

arXiv commentAccepted for presentation at SysML 2018

References in corpus (3)

• Intriguing properties of neural networks
• Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations
• Verification of Binarized Neural Networks via Inter-Neuron Factoring

Cited by in corpus (5)

• Testing and verification of neural-network-based safety-critical control software: A systematic literature review
• Automated Verification of Neural Networks: Advances, Challenges and Perspectives
• Enabling certification of verification-agnostic networks via memory-efficient semidefinite programming
• Pruning and Slicing Neural Networks using Formal Verification
• Towards Identifying and closing Gaps in Assurance of autonomous Road vehicleS -- a collection of Technical Notes Part 1