1 paper · 1 filter
Nicolas Angleraud, Antonia Karamolegkou, Benoît Sagot +1
Recent advances in visual language models (VLMs) have transformed end-to-end document understanding. However, their ability to interpret the complex layout semantics of historical…