This paper proposes a meta-learning algorithm that allows visual models to learn new concepts during inference without fine-tuning, similar to how large language models like ChatGPT learn new concepts. The approach outperforms state-of-the-art algorithms on several benchmarks without meta-training or fine-tuning.

https://arxiv.org/abs//2310.10971

YouTube: https://www.youtube.com/@ArxivPapers

TikTok: https://www.tiktok.com/@arxiv_papers

Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id1692476016

Spotify: https://podcasters.spotify.com/pod/show/arxiv-papers