Simple demonstration of separable convolutions

Keras w/ Tensorflow backend implementation for 3D channel-wise convolutions

Reimplementation of the PLS-Net architecture used for lung lobe segmentation in CT

Original code for "Exploiting Learned Symmetries in Group Equivariant Convolutions"

Classification between normal and pneumonia affected chest-X-ray images using deep residual learning along with separable convolutional network(CNN). This methodology involves efficient edge preservation and image contrast enhancement techniques for better classification of the X-ray images.

I Implemented some of the custom complex Convolutional Neural Network architecture using tensorow.keras Functional API.

Recognizing hindi letters written by tracing your hand in the air

A deep residual network implementing separable convolution to diagnose Pneumonia from CXR images.

Neural Style Transfer featuring styles of world's greatest artists.

A deep neural network developed following the residual learning and separable convolution paradigms to diagnose basal and squamous cell carcinoma using a subset of ISIC dataset.