TSRFormer: Table Structure Recognition with Transformers
This addresses the problem of accurately extracting structured data from table images for applications in document analysis and data digitization, representing a strong specific gain rather than a broad paradigm shift.
The paper tackles the problem of table structure recognition (TSR) for complex tables with geometrical distortions by proposing TSRFormer, which formulates line prediction as a regression task and uses a two-stage DETR-based approach with improvements for efficiency and accuracy. It achieves state-of-the-art performance on benchmark datasets like SciTSR, PubTabNet, and WTW, and demonstrates robustness on a challenging real-world dataset.
We present a new table structure recognition (TSR) approach, called TSRFormer, to robustly recognizing the structures of complex tables with geometrical distortions from various table images. Unlike previous methods, we formulate table separation line prediction as a line regression problem instead of an image segmentation problem and propose a new two-stage DETR based separator prediction approach, dubbed \textbf{Sep}arator \textbf{RE}gression \textbf{TR}ansformer (SepRETR), to predict separation lines from table images directly. To make the two-stage DETR framework work efficiently and effectively for the separation line prediction task, we propose two improvements: 1) A prior-enhanced matching strategy to solve the slow convergence issue of DETR; 2) A new cross attention module to sample features from a high-resolution convolutional feature map directly so that high localization accuracy is achieved with low computational cost. After separation line prediction, a simple relation network based cell merging module is used to recover spanning cells. With these new techniques, our TSRFormer achieves state-of-the-art performance on several benchmark datasets, including SciTSR, PubTabNet and WTW. Furthermore, we have validated the robustness of our approach to tables with complex structures, borderless cells, large blank spaces, empty or spanning cells as well as distorted or even curved shapes on a more challenging real-world in-house dataset.