The Alfeld and Van de Plas labs are currently offering a joint PhD position in instrumental and computational methods for multi-modal adaptive imaging, with a focus on imaging historical samples and cultural artefacts such as paintings. This position will lie at the interface between on the one hand spectral imaging instrumentation and hardware, and on the other hand mathematical engineering and machine learning.
Spectral imaging techniques, i.e. imaging modalities in which for each pixel a full spectrum is acquired, have changed the field of material characterization dramatically over the last decades, allowing for more in-depth characterization of complex samples and enabling a more exploratory approach to material investigation. One bottleneck in their application is that the techniques providing the richest insight are often relatively slow (i.e. they tend to require long acquisition times, or even may induce sample alteration). At the same time, relatively fast non-invasive techniques often provide only limited information. As a PhD candidate your task will be the development of adaptive imaging techniques for complex samples (a) by modeling the results of “slow” techniques based on “fast” techniques, using a combination of machine learning and physical simulations, and (b) by deriving from these predictions efficient image acquisition patterns with implemention on in-house built instruments. The candidate's tasks will be mainly oriented towards computational modeling and training on data from selected paintings, but will also include acquisition of new data from historical paintings to control the entire "chaîne opératoire", and preparing and performing scientific communication in written form (i.e. journal articles) and in oral form (i.e. speaking at conferences).
Plus d’informations :
[Website Delft University]