Deep learning models that are trained on histopathological images obtained\nfrom a single lab and/or scanner give poor inference performance on images\nobtained from another scanner/lab with a different staining protocol. In recent\nyears, there has been a good amount of research done for image stain\nnormalization to address this issue. In this work, we present a novel approach\nfor the stain normalization problem using fast neural style transfer coupled\nwith adversarial loss. We also propose a novel stain transfer generator network\nbased on High-Resolution Network (HRNet) which requires less training time and\ngives good generalization with few paired training images of reference stain\nand test stain. This approach has been tested on Whole Slide Images (WSIs)\nobtained from 8 different labs, where images from one lab were treated as a\nreference stain. A deep learning model was trained on this stain and the rest\nof the images were transferred to it using the corresponding stain transfer\ngenerator network. Experimentation suggests that this approach is able to\nsuccessfully perform stain normalization with good visual quality and provides\nbetter inference performance compared to not applying stain normalization.\n