We propose Selfdocseg, a self-supervised vision-based approach for document segmentation without requiring manual annotations...
S Maity, S Biswas, S Manna, A Banerjee, J Llados, S Bhattacharya, U Pal
International Conference on Document Analysis and Recognition, 2023 • 15 citations
We propose Selfdocseg, a self-supervised vision-based approach for document segmentation without requiring manual annotations...