FastMAE:Efficient masked autoencoder with offline tokenizerOA
FastMAE:Efficient masked autoencoder with offline tokenizer
Meng-Hao Guo;Chen Wang;Wei Liu;Shi-Min Hu
Department of Computer Science,Tsinghua University,Beijing 100084,ChinaUniversity of Pennsylvania,Philadelphia,PA 19104,USATencent Data Platform,Shenzhen 518057,ChinaDepartment of Computer Science,Tsinghua University,Beijing 100084,China
deep learningcomputer visionmasked image modeling(MIM)
deep learningcomputer visionmasked image modeling(MIM)
《计算可视媒体（英文）》 2025 (3)
483-496,14
This work was supported by the National Science and Technology Major Project(Grant No.2021ZD0112902),the National Natural Science Foundation of China(Grant Nos.623B2057 and 62220106003),Tsinghua University Initiative Scientific Research Program,and Tsinghua-Tencent Joint Laboratory for Internet Innovation Technology.The authors sincerely appreciate the dedicated effort and valuable feedback from the anonymous reviewers and editor,which significantly improved the manuscript.
评论