CVPR2025
FlexiDiT: Your Diffusion Transformer Can Easily Generate High-Quality Samples with Less Compute
Sotiris Anagnostidis, Gregor Bachmann, Yeongmin Kim, Jonas Kohler, Markos Georgopoulos, Artsiom Sanakoyeu, Yuming Du, Albert Pumarola, Ali K. Thabet, Edgar Schönfeld
摘要
agnostic to input and conditioning modalities. We show how our approach can be readily extended for video generation, where FlexiDiT models generate samples with up to 75% less compute without compromising performance.