2 papers
cs.CV2020
Validation and generalization of pixel-wise relevance in convolutional neural networks trained for face classification
Jñani Crawford, Eshed Margalit, Kalanit Grill-Spector +1
The increased use of convolutional neural networks for face recognition in science, governance, and broader society has created an acute need for methods that can show how these 'b…
cs.CV2018
The Effect of Learning Strategy versus Inherent Architecture Properties on the Ability of Convolutional Neural Networks to Develop Transformation Invariance
Megha Srivastava, Kalanit Grill-Spector
As object recognition becomes an increasingly common ML task, and recent research demonstrating CNNs vulnerability to attacks and small image perturbations necessitate fully unders…