by Ken Chiacchia | May 24, 2024 | AI, Bridges-2, Science Highlights
A prompt for Mexican dancers from an AI image generator produced the strange ballerinas at left; a new CMU-designed filter makes the image more appropriate and realistic, right. Image credit: Zhixuan Liu, Jean Oh, et al. 2024. SCoFT: Self-Contrastive Fine-Tuning for...
