1 paper
Xin Xie, Yu Guan, Zhuoxu Cui +2
By integrating the generative strengths of diffusion models with the representation capabilities of frequency-domain attention, DFAM effectively enhances reconstruction performance…