Mar 30, 2020 · We establish a benchmark suite consisting of different types of PDF document datasets that can be utilized for cross-domain DOD model training and evaluation.

Documents from the two domains may vary significantly in layout, language, and genre. We estab- lish a benchmark suite consisting of different types of PDF.

This repo provides a demo of how to use the proposed benchmark dataset. The code is heavily developed based on this repo, which is based on the faster-rcnn ...

This module is different from the previous methods [3, 15] that treat all region proposal equally, and instead focuses more on the hard-to-align pro- posals.

We establish a benchmark suite consisting of different types of PDF document datasets that can be utilized for cross-domain DOD model training and evaluation.

Mar 30, 2020 · A novel cross-domain DOD model is proposed which builds upon the standard detection model and addresses domain shifts by incorporating three ...

Recently, a new cross-domain Document Object Detection(DOD) benchmark was established in [24] to apply domain adaptation strategies for DOD to solve the domain ...

Project page for "Cross-Domain Document Object Detection: Benchmark Suite and Method, CVPR 2020". Python 45 5 · cvcZSL cvcZSL Public.

We establish a benchmark suite consisting of different types of PDF document datasets that can be utilized for cross-domain DOD model training and evaluation.

We establish a benchmark suite consisting of different types of PDF document datasets that can be utilized for cross-domain DOD model training and evaluation.