The paper introduces ITIT, a training paradigm for vision-language models that allows training on unpaired image and text data. ITIT achieves similar performance to models trained on paired data with significantly fewer data samples.

https://arxiv.org/abs//2310.03734

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