1 paper
Brandon Smock, Rohith Pesala, Robin Abraham
Benchmark datasets for table structure recognition (TSR) must be carefully processed to ensure they are annotated consistently. However, even if a dataset's annotations are self-co…