1
Learning Normalized Energy Models for Linear Inverse Problems
新能量模型破解生成扩散模型两大难题，实现线性逆问题无偏采样。
arXiv:2605.15487v1 Announce Type: new Abstract: Generative diffusion models can provide powerful prior probability models for inverse problems in imag…