method: PreSTU CC15M-SplitOCR B+B2022-09-20

Authors: Jihyung Kil, Soravit Changpinyo, Xi Chen, Hexiang Hu, Sebastian Goodman, Wei-Lun Chao, Radu Soricut

Affiliation: Google Research & The Ohio State University

Email: schangpi@google.com

Description: Baseline ViT-B/16 and mT5-Base using SplitOCR pre-training on CC15M (CC3M + CC12M)

Jihyung Kil, Soravit Changpinyo, Xi Chen, Hexiang Hu, Sebastian Goodman, Wei-Lun Chao, Radu Soricut. PreSTU: Pre-Training for Scene-Text Understanding. 2022.