Flamingo: a Visual Language Model for Few-Shot Learning

Authors: Alayrac et al. Year: 2022 ArXiv/Link: https://arxiv.org/abs/2204.14198

Summary

In-context multimodal learning model that learns from visual and textual examples, extending few-shot learning to vision.

Key Concepts

• In-context learning
• Multimodal examples
• Few-shot learning
• Visual reasoning
• Cross-modal understanding

Impact

Extended in-context learning to multimodal settings

Category

Video Understanding

---