Logographic Character Visual Pretraining via Semantic-based Contrastive Learning

作者：Daqian Shi, Wei Cao, Xiaoyu Zheng, Lida Shi, Xiaolei Diao, Cedric M John · 单位：Queen Mary University of London, Jilin University, King’s College London, University College London, Queen Mary University of London London UK, Jilin University Changchun China, King’s College London London UK, University College London London UK · 会议/期刊：arXiv preprint · 方向：cs.CV · 发布日期：2026-08-04