Abstract:
Inverse science comprises of decoding experimental observations of a given object to decipher various properties encoded within the object. In this talk, two of the most popular classes of inverse problem, namely computed tomography and super-resolution, will be presented. In the case of computed tomography, the inverse problem is formulated as the maximum a posteriori probability (MAP) estimation problem and a MAP cost function is iteratively minimized to deduce a most likely reconstructed result. Likewise, in the case of super-resolution the Convolutional Neural Networks (CNN) tool is used to scale-up low resolution images to their high resolution counterparts. The final assessment on the results obtained from our approaches on both of the aforementioned cases is made by the means of quantitative metrics such as the Root-Mean-Square Error (RMSE), the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) for synthetic as well as experimental datasets.