Micro expression recognition (MER)is a very challenging task as the\nexpression lives very short in nature and demands feature modeling with the\ninvolvement of both spatial and temporal dynamics. Existing MER systems exploit\nCNN networks to spot the significant features of minor muscle movements and\nsubtle changes. However, existing networks fail to establish a relationship\nbetween spatial features of facial appearance and temporal variations of facial\ndynamics. Thus, these networks were not able to effectively capture minute\nvariations and subtle changes in expressive regions. To address these issues,\nwe introduce an active imaging concept to segregate active changes in\nexpressive regions of a video into a single frame while preserving facial\nappearance information. Moreover, we propose a shallow CNN network: hybrid\nlocal receptive field based augmented learning network (OrigiNet) that\nefficiently learns significant features of the micro-expressions in a video. In\nthis paper, we propose a new refined rectified linear unit (RReLU), which\novercome the problem of vanishing gradient and dying ReLU. RReLU extends the\nrange of derivatives as compared to existing activation functions. The RReLU\nnot only injects a nonlinearity but also captures the true edges by imposing\nadditive and multiplicative property. Furthermore, we present an augmented\nfeature learning block to improve the learning capabilities of the network by\nembedding two parallel fully connected layers. The performance of proposed\nOrigiNet is evaluated by conducting leave one subject out experiments on four\ncomprehensive ME datasets. The experimental results demonstrate that OrigiNet\noutperformed state-of-the-art techniques with less computational complexity.\n