Researchers From The University Of York And Université Paris-saclay Introduce Deepknowledge For Generalisation-driven Deep Learning Testing

A recent study has introduced DeepKnowledge, a knowledge-driven test sufficiency criterion for DNN systems, to improve the consistency and reliability of deep neural networks in safety- and security-critical applications. This method analyzes the generalization behavior of the DNN model at the neuron level to ensure its performance in both training and unexpected data domain shifts.