Projects per year
Abstract
Spectral reconstruction (SR) algorithms attempt to recover hyperspectral information from RGB camera responses. Recently, the most common metric for evaluating the performance of SR algorithms is the Mean Relative Absolute Error (MRAE)—an ℓ _{1} relative error (also known as percentage error). Unsurprisingly, the leading algorithms based on Deep Neural Networks (DNN) are trained and tested using the MRAE metric. In contrast, the much simpler regressionbased methods (which actually can work tolerably well) are trained to optimize a generic Root Mean Square Error (RMSE) and then tested in MRAE. Another issue with the regression methods is—because in SR the linear systems are large and illposed—that they are necessarily solved using regularization. However, hitherto the regularization has been applied at a spectrum level, whereas in MRAE the errors are measured per wavelength (i.e., per spectral channel) and then averaged. The two aims of this paper are, first, to reformulate the simple regressions so that they minimize a relative error metric in training—we formulate both ℓ _{2} and ℓ _{1} relative error variants where the latter is MRAE—and, second, we adopt a perchannel regularization strategy. Together, our modifications to how the regressions are formulated and solved leads to up to a 14% increment in mean performance and up to 17% in worstcase performance (measured with MRAE). Importantly, our best result narrows the gap between the regression approaches and the leading DNN model to around 8% in mean accuracy.
Original language  English 

Article number  5586 
Journal  Sensors 
Volume  21 
Issue number  16 
DOIs  
Publication status  Published  19 Aug 2021 
Keywords
 Hyperspectral imaging
 Inverse problem
 Multispectral imaging
 Regression
 Regularization
 Spectral reconstruction
Projects
 1 Active

Established Career Fellowship
Finlayson, G. & Trollope, P.
Engineering and Physical Sciences Research Council
1/09/19 → 30/06/25
Project: Fellowship