Jianping Zhang

Replacing black-box image regression with an unrolled spectral unmixing objective physically disentangles structured sensor artifacts from true signals while boosting zero-shot transfer across complex noise regimes.

Decoupling structured sensor artifacts with a Deep Image Prior allows modular spectral unmixing to slash reconstruction error by up to 69.5% without corrupting physical abundance estimates under flawed endmember priors.

Sparse pixel-level supervision can yield segmentation results on par with fully supervised methods, revolutionizing the efficiency of training deep learning models.

Achieving accurate material decomposition in sparse-view DECT could revolutionize medical imaging by enabling safer, lower-radiation scans without sacrificing detail.

A novel approach that combines attention mechanisms with variational image registration achieves superior performance while enhancing explainability and modularity in medical imaging.