赞
踩
题目: DocPedia: Unleashing the Power of Large Multimodal Model in the Frequency Domain for Versatile Document Understanding
机构:中科大,字节
论文: https://arxiv.org/pdf/2311.11810.pdf
代码: 未开源
任务: OCR-free document understanding(感知与认知)
特点: 非像素空间,而是频域空间处理
前置工作:LLaVAR [55], mPLUG-DocOwl [51], and UniDoc [9],Donut [18], KOSMOS-2.5 [31], and Pix2Struct [20],UReader [52]
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。