no code implementations • 12 Mar 2024 • Michael Ogezi, Ning Shi
In text-to-image generation, using negative prompts, which describe undesirable image characteristics, can significantly boost image quality.
1 code implementation • 24 Jun 2023 • Michael Ogezi, Bradley Hauer, Talgat Omarov, Ning Shi, Grzegorz Kondrak
We present a novel algorithm that leverages glosses retrieved from BabelNet, in combination with text and image encoders.
no code implementations • 5 Jun 2023 • Michael Ogezi, Bradley Hauer, Grzegorz Kondrak
With both ideas, we demonstrate that V-GLOSS improves visual descriptions and achieves strong results in the zero-shot setting on general and fine-grained image-classification datasets, including ImageNet, STL-10, FGVC Aircraft, and Flowers 102.