Document images are a ubiquitous source of data where the text is organi...
In this paper, we introduce a novel concept of user-entity differential
...
Cross-Domain Detection (XDD) aims to train an object detector using labe...
Decomposing images of document pages into high-level semantic regions (e...