Zero-shot Image-to-Image Translation — mentioned by imranq

Tue, 14 Feb 2023 15:56:51 GMT

Actually this is wrong, after reading the paper: https://arxiv.org/pdf/2302.03027.pdf The main idea is to specify an edit direction which is a broad statement of what you want to do. The model then uses gpt-3 to generate text for that direction and then it calculates embeddings with CLIP and then the mean difference. The difference is used to guide the noise to make the final change. One pitfall seems like you can't specify very customized changes, but very cool approach

Paper and discussion history. A mention is not an endorsement.

Zero-shot Image-to-Image Translation · HackerPapers discussions

Zero-shot Image-to-Image Translation — mentioned by imranq