Nuclear Diffusion Models for Low-Rank Background Suppression in Videos
This work addresses video restoration challenges in medical imaging, such as cardiac ultrasound, by combining model-based and deep learning approaches, though it is incremental as it builds on existing low-rank and diffusion methods.
The paper tackled the problem of suppressing low-rank background noise in videos, particularly for cardiac ultrasound dehazing, by proposing a hybrid framework that integrates low-rank temporal modeling with diffusion posterior sampling. The result was improved dehazing performance, with better contrast enhancement (gCNR) and signal preservation (KS statistic) compared to traditional RPCA methods.
Video sequences often contain structured noise and background artifacts that obscure dynamic content, posing challenges for accurate analysis and restoration. Robust principal component methods address this by decomposing data into low-rank and sparse components. Still, the sparsity assumption often fails to capture the rich variability present in real video data. To overcome this limitation, a hybrid framework that integrates low-rank temporal modeling with diffusion posterior sampling is proposed. The proposed method, Nuclear Diffusion, is evaluated on a real-world medical imaging problem, namely cardiac ultrasound dehazing, and demonstrates improved dehazing performance compared to traditional RPCA concerning contrast enhancement (gCNR) and signal preservation (KS statistic). These results highlight the potential of combining model-based temporal models with deep generative priors for high-fidelity video restoration.