Fetching the paper…

LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking · Around