TextFormer: A Query-based End-to-End Text Spotter with Mixed Supervision

arXiv:2306.03377v2 Announce Type: replace
Abstract: End-to-end text spotting is a vital computer vision task that aims to integrate scene text detection and recognition into a unified framework. Typical methods heavily rely on Region-of-Interest (RoI) operations to extract local features and complex post-processing steps to produce final predictions. To address these limitations, we propose TextFormer, a query-based end-to-end text spotter with Transformer architecture. Specifically, using query embedding per text instance, TextFormer builds upon an image encoder and a text decoder to …

abstract arxiv computer computer vision cs.cl cs.cv detection extract features framework limitations mixed operations post-processing predictions processing query recognition roi supervision text type vision vital