Scientist for machine learning synchrotron applications

The Extremely Brilliant Source upgrade of the ESRF has led to a significant increase in the acquisition speed and the amount of data which must be processed and analysed by the end users. It is therefore essential to provide advanced tools to enhance the acquisition and data analysis strategies, and provide online (live during experiments) feedback to the ESRF users.

We are looking to recruit two 5-year scientists to work on machine learning developments in two main areas: firstly, for tomography, e.g. for applications like denoising, super-resolution, segmentation or improved data collection. And secondly for scattering/diffraction applications, notably (but not exclusively) for structural biology, where the standardisation of techniques makes it a prime target for smarter automation to improve the data quality and the acquisition efficiency.

The work will take place in the Algorithms & scientific Data Analysis group, in close collaboration with ESRF beamlines. Further information about the post can be obtained from Vincent Favre-Nicolin (favre@esrf.fr).

Your missions include:

• The development and deployment of innovative machine-learning based algorithms for data acquisition and online data processing, in the fields of scattering/diffraction (notably structural biology) and tomography

• Conduct an original research project exploiting machine learning for synchrotron applications with ESRF beamlines scientists

• Work with scientists and engineers at ESRF and with partner institutions to expand and disseminate knowledge about the use of ML for synchrotron applications