The project is a research effort that explores the susceptibility of deep neural networks, particularly those used for image classification, to minimal input perturbations. It demonstrates the ability to manipulate a single pixel in an image, causing the neural network to misclassify the image.
sachin576123 / one-pixel-attack-for-fooling-depp-neural-network Goto Github PK
View Code? Open in Web Editor NEWThe project is a research effort that explores the susceptibility of deep neural networks, particularly those used for image classification, to minimal input perturbations. It demonstrates the ability to manipulate a single pixel in an image, causing the neural network to misclassify the image.