Back to Reading Room

LA BIBLIOTHÈQUE · READING NOTE

Learning Transferable Visual Models From Natural Language Supervision

Radford et al.

To Read · 想看

AIMultimodalPapers

CLIP, Contrastive Learning, Vision-Language, Zero-Shot

CLIP 通过大规模图文对进行对比学习,将视觉与自然语言映射到共享表示空间,对后来的多模态模型和 Text-to-Image 系统影响深远。