Table Structure Recognition is an essential part of end-to-end tabular data\nextraction in document images. The recent success of deep learning model\narchitectures in computer vision remains to be non-reflective in table\nstructure recognition, largely because extensive datasets for this domain are\nstill unavailable while labeling new data is expensive and time-consuming.\nTraditionally, in computer vision, these challenges are addressed by standard\naugmentation techniques that are based on image transformations like color\njittering and random cropping. As demonstrated by our experiments, these\ntechniques are not effective for the task of table structure recognition. In\nthis paper, we propose TabAug, a re-imagined Data Augmentation technique that\nproduces structural changes in table images through replication and deletion of\nrows and columns. It also consists of a data-driven probabilistic model that\nallows control over the augmentation process. To demonstrate the efficacy of\nour approach, we perform experimentation on ICDAR 2013 dataset where our\napproach shows consistent improvements in all aspects of the evaluation\nmetrics, with cell-level correct detections improving from 92.16% to 96.11%\nover the baseline.\n