Task-adaptive eigenvector-based projection (EBP) transform for compressed sensing: A case study of spectroscopic profiling sensor
Task-adaptive eigenvector-based projection (EBP) transform for compressed sensing: A case study of spectroscopic profiling sensor
复制标题
用于压缩感知的任务自适应基于特征向量的投影（EBP）变换：光谱分析传感器的案例研究
DOI：
10.1002/ansa.202100018
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Qin Xiaolin
中科院分区：
文献类型：
--
作者：
Zhang Yinsheng;Wang Haiyan;Cheng Yongbo;Qin Xiaolin
压缩感知（CS）理论要求在某种变换下信号是稀疏的。对于大多数信号（例如，语音和照片），非自适应变换基，如离散余弦变换（DCT），离散傅立叶变换（DFT）和Walsh - Hadamard变换（WHT），可以满足这一要求，并表现得很好。然而，这些非自适应变换的一个限制是我们不能利用特定领域的知识来提高CS效率。本研究提出一种任务自适应特征向量投影（EBP）变换。EBP基具有与主成分加载矩阵等效的作用，可以在潜在空间中生成稀疏表示。在拉曼光谱分析案例研究中，EBP表现出比非自适应对应物更好的性能。在1% CS采样比(k)下，DCT、DFT、WHT和EBP的重建相对均方误差分别为0.33、0.68、0.32和0.00。在一定范围内，EBP比非自适应的重建质量要好得多。对于特定的域任务，EBP可以显著降低CS采样率，降低总体测量成本。
The compressed sensing (CS) theory requires the signal to be sparse under some transform. For most signals (e.g., speech and photos), the non‐adaptive transform bases, such as discrete cosine transform (DCT), discrete Fourier transform (DFT), and Walsh‐Hadamard transform (WHT), can meet this requirement and perform quite well. However, one limitation of these non‐adaptive transforms is that we cannot leverage domain‐specific knowledge to improve CS efficiency. This study presents a task‐adaptive eigenvector‐based projection (EBP) transform. The EBP basis has an equivalent effect of the principal component loading matrix and can generate a sparse representation in the latent space. In a Raman spectroscopic profiling case study, EBP demonstrates better performance than its non‐adaptive counterparts. At the 1% CS sampling ratio (k), the reconstruction relative mean square errors of DCT, DFT, WHT and EBP are 0.33, 0.68, 0.32, and 0.00, respectively. At a fixedk, EBP achieves much better reconstruction quality than the non‐adaptive counterparts. For specific domain tasks, EBP can significantly lower the CS sampling ratio and reduce the overall measurement cost.