1 paper
Jiatong Xia, Zicheng Duan, Anton van den Hengel +1
Recent progress in 3D generation has been driven largely by models conditioned on images or text, while readily available 3D priors are still underused. In many real-world scenario…