CLIP-Driven Fine-Grained Text-Image Person Re-Identification | IEEE Journals & Magazine | IEEE Xplore