HiRAE: Hierarchical Representation Autoencoding with Residual Budgets
Pretrained visual representations support image generation, but may not fully preserve the fine-grained details needed for faithful reconstruction. Meanwhile, intermediate encoder layers contain complementary visual details, but learning to fuse them for reconstruction can produce a latent distribution that is difficul...
Xuan-Yu Zhu, Yan Bai, Yang Shi et al.
· 0 citations