Abstract Compressed sensing (CS) is a pioneering sub-Nyquist sampling technique that reconstructs signals from far fewer measurements than traditional acquisition schemes. Considerable amount of research has been carried out over the years to develop powerful algorithms that can accomplish optimal recovery. However, they vary considerably in their ease of implementation, speed of recovery and noise resilience. This work presents the Sparsity Independent Regularized Pursuit (SIRP) which achieves an admirable trade-off between these key features. Further, it requires no prior knowledge of exact sparsity level and possesses a regular structure, making it amenable to low cost hardware solutions. Experimental investigations reveal the competitiveness of SIRP with existing state-of-art in regard to successful recovery and significant speed-up. The algorithm also attains superior results on a considerably large image recovery problem, which demonstrates its suitability to real world applications.

中文翻译：

---

使用稀疏独立正则化追踪从压缩测量中恢复

摘要 压缩感知 (CS) 是一种开创性的亚奈奎斯特采样技术，与传统的采集方案相比，它可以从更少的测量中重建信号。多年来，已经进行了大量研究以开发可以实现最佳恢复的强大算法。然而，它们在实施的难易性、恢复速度和抗噪能力方面差异很大。这项工作提出了稀疏独立正则化追踪（SIRP），它在这些关键特征之间实现了令人钦佩的权衡。此外，它不需要精确稀疏级别的先验知识，并且具有规则结构，使其适用于低成本的硬件解决方案。实验研究揭示了 SIRP 与现有技术在成功恢复和显着加速方面的竞争力。