Annonce

Les commentaires sont clos.

stage recherche M2 en imagerie biologique

2 Novembre 2021


Catégorie : Stagiaire


Super-resolution fluorescent microscopy using Variational AutoEncoder (VAEs) priors
M2 internship proposal for spring 2022 (Duration: 5/6 months) within the 3IA Côte d'Azur project.
Supervisors: Laure Blanc-F´eraud (blancf@i3s.unice.fr), Luca Calatroni (calatroni@i3s.unice.fr), Sebastien Schaub (sebastien.schaub@imev-mer.fr)
Host institution: MORPHEME research group (INRIA, CNRS, I3S, Sophia-Antipolis, France).

 

Context
Conventional optical microscopy techniques, as confocal microscopes, are widely used in biology for cellular and sub-cellular structures investigation. However, their spatial resolution is limited by the light diffraction phenomena and it is typically around 200nm in the transverse plane and 400nm in the optical axis. Over the recent years, several super-resolution techniques have been developed to overcome this drawback. Popular techniques, such that, for example, SMLM (Single Molecule Localization Microscopy) are based on the use photoactivable molecules, others on structured illumination or on the analysis of the stochastic fluctuation of molecules (such as SOFI) and many more [1]. In the Morpheme group we have developed advanced algorithms for SMLM and for the analysis of SOFI-type images based on sparse optimization reconstruction methods, see, e.g. [2, 3]. The super-resolved image reconstruction is formalized in mathematical terms as an inverse problem which is regularized by introducing a sparsity-promoting penalization. A different and increasingly popular class of methods producing outstanding results in many applied fields is based on the use of modern generative learning approaches. Among them, Variational Autoencoders (VAEs) [4] have attracted the attention of several researchers working in the field of inverse problems due to their ability of combining variational inference approaches with the ability of neural networks to learn unknown posterior distributions distributions. Their use in the field of image microscopy, however, remains limited

Internship objectives
The purpose of this internship is to develop an inverse super-resolution method based on the use of VAEs, inspired, e.g., from recent works in compressed sensing and general inverse problems [5, 6]. The main idea consists in developing a super-resolved VAE-based reconstruction method for studying molecule fluctuations.
Possible difficulties arising in the optimisation steps may be mitigated by performing alternating minimisation both in the image and in the latent space [6] and improved results can be achieved, e.g., by using improved architectures [7]. Such approaches are of utmost interest for the biological community as they are harmless for the sample, they do not need any specific microscope and maintain adequate time resolution. The performance of this approach will be compared with the model-driven (e.g., COL0RME [3]) and data-driven (e.g., GANs) approaches already developed in the Morpheme research group on synthetic and real data.

Candidate profile
Second year of Master degree in computer science, applied mathematics, data science with background in image processing, imaging inverse problems, deep learning and optimisation. Good coding skills for numerical simulation (Pytorch, Python, MATLAB, ...). A general interest in health and biology is welcome.

Practical information
MORPHEME research team is a joint research group between INRIA Sophia Antipolis M´editerran´ee., I3S Lab (Universit´e Cˆote d’Azur and CNRS).
Remuneration: internship gratification (approximately 550 euros/month) and possible discounts for nearby accommodation facilities (CIV).

 

Application procedure
Please send your CV, motivation letter, marks of the last two years of study and the name and e-mail address of a contact for recommendation to Laure Blanc-F´eraud (blancf@i3s.unice.fr), Luca Calatroni (calatroni@i3s.unice.fr).

References
[1] L. Schermelleh, A. Ferrand, T. Huser, C. Eggeling, M. Sauer, O. Biehlmaier, and G. P. C. Drummen,
“Super-resolution microscopy demystified,” Nature Cell Biology, vol. 21, no. 1, pp. 72–84, 2019.
[2] S. Gazagnes, E. Soubies, and L. Blanc-F´eraud, “High density molecule localization for super-resolution
microscopy using CEL0 based sparse approximation,” in 2017 IEEE 14th International Symposium on
Biomedical Imaging (ISBI 2017), pp. 28–31, 2017.
[3] V. Stergiopoulou, L. Calatroni, J. H. de Morais Goulart, S. Schaub, and L. Blanc-F´eraud, “Col0rme:
Super-resolution microscopy based on sparse blinking fluorophore localization and intensity estimation,”
2021. arXiv preprint: https://arxiv.org/abs/2108.07095.
[4] D. P. Kingma and M. Welling, “Auto-Encoding Variational Bayes,” in 2nd International Conference
on Learning Representations, ICLR 2014, Banff, AB, Canada, April 14-16, 2014, Conference Track
Proceedings, 2014.
[5] A. Bora, A. Jalal, E. Price, and A. G. Dimakis, “Compressed sensing using generative models,” in Proceedings
of the 34th International Conference on Machine Learning - Volume 70, ICML’17, p. 537–546,
JMLR.org, 2017.
[6] M. Gonz´alez, A. Almansa, and P. Tan, “Solving inverse problems by joint posterior maximization with
autoencoding prior,” 2021. arXiv preprint: https://arxiv.org/abs/2103.01648.
[7] A. Vahdat and J. Kautz, “Nvae: A deep hierarchical variational autoencoder,” ArXiv,
vol. abs/2007.03898, 2020.