A Transformer-based object detection model pre-trained on FinTabNet for recognizing table structures, including row and column detection, in financial and tabular documents.
Table Transformer (TATR) - FinTabNet Table Structure Recognition is a Transformer-based object detection model trained on the FinTabNet dataset. It is designed for detecting and recognizing table structures, such as rows, columns, and table boundaries, in financial and structured tabular documents. Built on the DETR (DEtection TRansformer) framework, the model applies a "normalize before" approach, where layer normalization occurs before self- and cross-attention mechanisms. This model is particularly useful for table parsing, financial document processing, and extracting structured data from scanned reports and spreadsheets.
MIT
Microsoft
object detection
N.A.
Open
Sector Agnostic
12/03/25 06:34:58
0
MIT
© 2026 - Copyright AIKosh. All rights reserved. This portal is developed by National e-Governance Division for AIKosh mission.