Robust fine-tuning of zero-shot models
CLIP: Connecting Text and Images: We’re introducing a neural network called CLIP which efficiently learns visual concepts from natural language supervision. CLIP can be applied to any visual classification benchmark by simply providing the names of the visual categories to be recognized, similar to the ‘zero-shot’ capabilities of GPT-2 and GPT-3
https://arxiv.org/abs/2102.05918
https://arxiv.org/abs/2010.11929
https://arxiv.org/abs/1707.02968
ImageNet Large Scale Visual Recognition Challenge