DeepMind Claims Image Captioner Alone Is Surprisingly Powerful then Previous Believed, Competing with CLIP

In a new paper Image Captioners Are Scalable Vision Learners Too, a DeepMind research team presents CapPa, a image captioning based pretraining strategy that and can compete CLIP and exhibit favorable model and data scaling properties, verifying that a plain image captioning can be a competitive pretraining strategy for vision backbones.

ai artificial intelligence captioning clip computer vision & graphics contrastive model data deepmind deepmind research deep-neural-networks image machine learning machine learning & data science ml paper research research team scalable scaling strategy team technology vision vision-transformer