Generating photorealistic images of human subjects in any unseen pose have\ncrucial applications in generating a complete appearance model of the subject.\nHowever, from a computer vision perspective, this task becomes significantly\nchallenging due to the inability of modelling the data distribution conditioned\non pose. Existing works use a complicated pose transformation model with\nvarious additional features such as foreground segmentation, human body parsing\netc. to achieve robustness that leads to computational overhead. In this work,\nwe propose a simple yet effective pose transformation GAN by utilizing the\nResidual Learning method without any additional feature learning to generate a\ngiven human image in any arbitrary pose. Using effective data augmentation\ntechniques and cleverly tuning the model, we achieve robustness in terms of\nillumination, occlusion, distortion and scale. We present a detailed study,\nboth qualitative and quantitative, to demonstrate the superiority of our model\nover the existing methods on two large datasets.\n