The objective of person re-identification (re-ID) is to retrieve a person's\nimages from an image gallery, given a single instance of the person of\ninterest. Despite several advancements, learning discriminative\nidentity-sensitive and viewpoint invariant features for robust Person\nRe-identification is a major challenge owing to the large pose variation of\nhumans. This paper proposes a re-ID pipeline that utilizes the image generation\ncapability of Generative Adversarial Networks combined with pose clustering and\nfeature fusion to achieve pose invariant feature learning. The objective is to\nmodel a given person under different viewpoints and large pose changes and\nextract the most discriminative features from all the appearances. The pose\ntransformational GAN (pt-GAN) module is trained to generate a person's image in\nany given pose. In order to identify the most significant poses for\ndiscriminative feature extraction, a Pose Clustering module is proposed. The\ngiven instance of the person is modelled in varying poses and these features\nare effectively combined through the Feature Fusion Network. The final re-ID\nmodel consisting of these 3 sub-blocks, alleviates the pose dependence in\nperson re-ID. Also, The proposed model is robust to occlusion, scale, rotation\nand illumination, providing a framework for viewpoint invariant feature\nlearning. The proposed method outperforms the state-of-the-art GAN based models\nin 4 benchmark datasets. It also surpasses the state-of-the-art models that\nreport higher re-ID accuracy in terms of improvement over baseline.\n