From Generality to Specificity: Prior-Driven Optimal Sparse Transformation in Compressed Sensing

TL;DR

This paper presents the POST framework that unifies generalization and specificity in sparse transforms for compressed sensing, introducing HOT as an adaptive domain that improves performance across various signal types and applications.

Contribution

The paper proposes the POST framework and derives HOT, a novel sparse transform domain that adaptively balances generalization and specificity in compressed sensing.

Findings

HOT achieves superior sparse representations for real and complex signals.

HOT improves performance in audio sensing, 5G channel estimation, and image compression.

HOT enhances multiple compressed sensing algorithms with negligible computational cost.

Abstract

This paper introduces a new paradigm for sparse transformation: the Prior-to-Posterior Sparse Transform (POST) framework, designed to overcome long-standing limitation on generalization and specificity in classical sparse transforms for compressed sensing. POST systematically unifies the generalization capacity of any existing transform domains with the specificity of reference knowledge, enabling flexible adaptation to diverse signal characteristics. Within this framework, we derive an explicit sparse transform domain termed HOT, which adaptively handles both real and complex-valued signals. We theoretically establish HOT's sparse representation properties under single and multiple reference settings, demonstrating its ability to preserve generalization while enhancing specificity even under weak reference information. Extensive experiments confirm that HOT delivers substantial meta-gains across audio sensing, 5G channel estimation, and image compression tasks, consistently boosting multiple compressed sensing algorithms under diverse multimodal settings with negligible computational overhead.