Entrepôts, Représentation et Ingénierie des Connaissances
Seminars of the ERIC lab
For questions about the seminars, please contact Fadi Bentayeb


by Year
by Type
by Topic

(1) Presentation(s)


Mon 01/14/19 11:00 K71, Buiding K, RdC

GANAYE Pierre-antoine (Creatis - INSA Lyon)
Semi-supervised learning for segmentation under semantic constraint


Image segmentation based on convolutional neural networks
is proving to be a powerful and efficient solution for medical applications.
However, the lack of annotated data, presence of artifacts and variability
in appearance can still result in inconsistencies during the inference. We
choose to take advantage of the invariant nature of anatomical struc-
tures, by enforcing a semantic constraint to improve the robustness of
the segmentation. The proposed solution is applied on a brain struc-
tures segmentation task, where the output of the network is constrained
to satisfy a known adjacency graph of the brain regions. This criteria
is introduced during the training through an original penalization loss
named NonAdjLoss. With the help of a new metric, we show that the
proposed approach significantly reduces abnormalities produced during
the segmentation. Additionally, we demonstrate that our framework can
be used in a semi-supervised way, opening a path to better generalization
to unseen data.

For more information please contact Cugliari J.