CLIP4STR: A Simple Baseline for Scene Text Recognition With Pre-Trained Vision-Language Model | IEEE Journals & Magazine | IEEE Xplore