Transformational Measures

Transformational measures allow to quantify the relationship between transformations of the inputs to neural networks and their outputs or intermediate representations.

In particular

Measuring transformation (in)variances in Neural Networks

Measures of invariance to transformations such as rotations, scaling and translation for Deep Neural Networks

Transformational Measures Experiments

Experiments to measure invariance of various well-known CNN architectures such as VGG, ResNet and AllConvolutional, on datasets CIFAR10 and MNIST.

Transformational Measures for Neural Networks - PyTorch Library

A library to compute easily compute Transformation Measures for PyTorch Models.

Transformational Measures for Neural Networks - Keras/TensorFlow Library

A library to compute easily compute Transformation Measures for Keras/Tensorflow Models.