[2602.23447] SALIENT: Frequency-Aware Paired Diffusion for Controllable Long-Tail CT Detection

Electrical Engineering and Systems Science > Image and Video Processing arXiv:2602.23447 (eess) [Submitted on 26 Feb 2026] Title:SALIENT: Frequency-Aware Paired Diffusion for Controllable Long-Tail CT Detection Authors:Yifan Li, Mehrdad Salimitari, Taiyu Zhang, Guang Li, David Dreizin View a PDF of the paper titled SALIENT: Frequency-Aware Paired Diffusion for Controllable Long-Tail CT Detection, by Yifan Li and 4 other authors View PDF HTML (experimental) Abstract:Detection of rare lesions in whole-body CT is fundamentally limited by extreme class imbalance and low target-to-volume ratios, producing precision collapse despite high AUROC. Synthetic augmentation with diffusion models offers promise, yet pixel-space diffusion is computationally expensive, and existing mask-conditioned approaches lack controllable attribute-level regulation and paired supervision for accountable training. We introduce SALIENT, a mask-conditioned wavelet-domain diffusion framework that synthesizes paired lesion-masking volumes for controllable CT augmentation under long-tail regimes. Instead of denoising in pixel space, SALIENT performs structured diffusion over discrete wavelet coefficients, explicitly separating low-frequency brightness from high-frequency structural detail. Learnable frequency-aware objectives disentangle target and background attributes (structure, contrast, edge fidelity), enabling interpretable and stable optimization. A 3D VAE generates diverse volumetric lesion masks, ...