Doctoral Fellowship In Space Science And Technology

Position : 3-year doctoral fellowship in Space Science and Technology.

Title : Deep Learning techniques for inverse problem in imaging (CUP E66E24000000001)

Doctoral site : University of Brescia

Mobility abroad : compulsory, minimum 6 months

Description : Inverse problems deal with the reconstruction of an unknown signal, image, or multi-dimensional volume from a set of observations, which are generally the result of a non-invertible forward process. A wide range of scenarios fall within this definition, such as deconvolution, image deblurring, inpainting, and so on. As inverse problems are usually ill-posed, in general, it is not possible to find a unique solution that describes the observation, unless some prior knowledge about the data is available. Traditional approaches, based on the minimization of a cost function and a regularizer taking into account possible prior knowledge on data, are found to underperform with respect to recent deep learning techniques. The goal of the project is to explore the potentiality of some modern deep learning techniques in selected imaging cases, such as muography and medical imaging, which are relevant to science and industry.

The student will be based in Brescia (Italy) but at least 6 months (up to 12 months) will be spent abroad. The application is online (see link below) and the deadline is July 16th, 2024 hrs. 4.00 PM (Italian time). The selected candidate will start on November 1 (2024) .

To apply : Click here

#J-18808-Ljbffr