Deep learning has become the state-of-the-art approach for the segmentation of biomedical images.
In the case of histology, staining is often used to highlight anatomical structures. For biological or
technical reasons, staining techniques often produce variable results. Data available at a given
moment is therefore not always representative of future acquisitions.
The aim of this internship is to propose different solutions to this problem, such as image
augmentation, backpropagation for adaptation , image to image translation or generative
adversarial networks .
Candidates should be familiar with machine learning and image processing. Experience with
Python for programming would be a bonus. The internship could lead to a Ph.D. position.
Contact : firstname.lastname@example.org, email@example.com
 Sankaranarayanan, Swami, et al. "Generate to adapt: Aligning domains using generative
adversarial networks." CVPR 2018.
 Motiian, Saeid, et al. "Few-shot adversarial domain adaptation." Advances in Neural
Information Processing Systems (NIPS), 2017.
 Murez, Zak, et al. "Image to Image Translation for Domain Adaptation." Proceedings of the
IEEE Conference on Computer Vision and Pattern Recognition. 2018.
 Hosseini-Asl, E., Zhou, Y., Xiong, C., & Socher, R.. Augmented Cyclic Adversarial Learning for
Domain Adaptation. ICML 2018
 Ganin, Yaroslav, and Victor Lempitsky. "Unsupervised domain adaptation by
backpropagation." ICML 2015
(c) GdR 720 ISIS - CNRS - 2011-2018.