Sample-Specific Output Constraints for Neural Networks
This addresses the need for safer and more interpretable neural networks in safety-critical domains like autonomous vehicles, though it appears incremental as it modifies existing architectures.
The authors tackled the problem of neural networks being black boxes by proposing ConstraintNet, a neural network that constrains outputs via an additional input to exclude unintended or hazardous predictions, demonstrating its application in facial landmark prediction and a safety-critical vehicle controller.
Neural networks reach state-of-the-art performance in a variety of learning tasks. However, a lack of understanding the decision making process yields to an appearance as black box. We address this and propose ConstraintNet, a neural network with the capability to constrain the output space in each forward pass via an additional input. The prediction of ConstraintNet is proven within the specified domain. This enables ConstraintNet to exclude unintended or even hazardous outputs explicitly whereas the final prediction is still learned from data. We focus on constraints in form of convex polytopes and show the generalization to further classes of constraints. ConstraintNet can be constructed easily by modifying existing neural network architectures. We highlight that ConstraintNet is end-to-end trainable with no overhead in the forward and backward pass. For illustration purposes, we model ConstraintNet by modifying a CNN and construct constraints for facial landmark prediction tasks. Furthermore, we demonstrate the application to a follow object controller for vehicles as a safety-critical application. We submitted an approach and system for the generation of safety-critical outputs of an entity based on ConstraintNet at the German Patent and Trademark Office with the official registration mark DE10 2019 119 739.