1 papers with code • 0 benchmarks • 0 datasets

This task has no description! Would you like to contribute one?

Most implemented papers

GeoLayoutLM: Geometric Pre-training for Visual Information Extraction

alibabaresearch/advancedliteratemachinery CVPR 2023

Additionally, novel relation heads, which are pre-trained by the geometric pre-training tasks and fine-tuned for RE, are elaborately designed to enrich and enhance the feature representation.