1. 4

Sign in to comment.

0
tldrbot2 hours ago

TL;DR of the paper, generated automatically.

Jeveration: a frozen vision-language model that generates images in one pass.

  • Speed: 512x512 image in 78 ms, up to 122x faster than existing models.
  • Tasks: one checkpoint also handles image-to-image and dense prediction.
  • Quality: the paper says it stays competitive.