{"id":177,"date":"2022-09-16T11:23:04","date_gmt":"2022-09-16T11:23:04","guid":{"rendered":"https:\/\/blog.metu.edu.tr\/figeno\/?page_id=177"},"modified":"2023-11-01T08:51:55","modified_gmt":"2023-11-01T08:51:55","slug":"1001project","status":"publish","type":"page","link":"https:\/\/blog.metu.edu.tr\/figeno\/projects\/1001project\/","title":{"rendered":""},"content":{"rendered":"\n

Development of Deep Learning-based Approaches for Solving Inverse Problems in Imaging and Comparative Performance Evaluation<\/mark><\/h1>\n\n\n\n

Project Coordinator:<\/strong> Assoc. Prof. Figen S. Oktem<\/a><\/p>\n\n\n\n

Project Type:<\/strong> T\u00dcB\u0130TAK 1001 Scientific and Technological Research Projects Funding Program<\/p>\n\n\n\n

Project Budget:<\/strong> 869.686 TL<\/p>\n\n\n\n

Project Duration:<\/strong> 36 Months<\/p>\n\n\n\n

Project Start Date:<\/strong> 15 March 2021<\/p>\n\n\n\n

Funded Personnel:<\/strong> 1 PhD student (full-time), 2 MSc students (full-time)<\/p>\n\n\n\n

Project Summary:<\/strong> <\/p>\n\n\n\n

Computational imaging, a rapidly evolving interdisciplinary field, enables new forms of visual information in various applications in natural sciences. In a computational imaging system, an inverse problem has to be solved to reconstruct an image from the acquired raw data. For the solution of these high-dimensional inverse problems, commonly used fast direct inversion methods are not robust to noise. On the other hand, regularization-based methods can offer better reconstruction quality but with higher computational cost.<\/p>\n\n\n\n

Recently deep-learning based approaches have been developed to achieve high accuracy with fast reconstruction. In this project, by using a general framework for various inverse problems in imaging, we will develop deep learning-based fast techniques that enable unprecedented reconstruction quality. Moreover, the advantages and disadvantages of the developed techniques will be investigated not only through numerical simulations but also experimentally. For this purpose, inverse problems will be studied by considering the following three general categories separately: a) two-dimensional (2D) linear problems, b) two-dimensional nonlinear problems, c) three- or more-dimensional linear problems. Because the existing deep learning-based methods are mostly developed for two-dimensional linear problems, this project focuses on the development of new methods mostly for the two-dimensional nonlinear problems, and three or more-dimensional linear problems. Likewise, deep learning-based techniques will be studied by considering the following general categories: i) learned direct reconstruction,<\/em> ii) reconstruction with learned prior<\/em>, iii) learned iterative reconstruction<\/em>. Different type of techniques will be developed separately, and then will be compared with each other and commonly used analytical inversion methods. By demonstrating the pros and cons of each approach, it is expected to understand which approaches perform better for different inverse problems and measurement settings. <\/p>\n\n\n\n

Project Team Members:<\/strong> <\/p>\n\n\n\n