1 paper
Chang Xie, Chenyi Zhuang, Pan Gao
Advanced diffusion models have made notable progress in text-to-image compositional generation. However, it is still a challenge for existing models to achieve text-image alignment…