S
Soufiane Belharbi
Researcher at École de technologie supérieure
Publications - 23
Citations - 257
Soufiane Belharbi is an academic researcher from École de technologie supérieure. The author has contributed to research in topics: Computer science & Deep learning. The author has an hindex of 7, co-authored 23 publications receiving 152 citations. Previous affiliations of Soufiane Belharbi include Intelligence and National Security Alliance.
Papers
More filters
Journal ArticleDOI
Spotting L3 slice in CT scans using deep convolutional network and transfer learning.
Soufiane Belharbi,Clément Chatelain,Romain Hérault,Sébastien Adam,Sébastien Thureau,Mathieu Chastan,Romain Modzelewski +6 more
TL;DR: A complete automated system for spotting a particular slice in a complete 3D Computed Tomography exam (CT scan) that relies on an original machine learning regression approach and is applied to the detection of the third lumbar vertebra (L3) slice that has been found to be representative to the whole body composition.
Posted Content
Deep weakly-supervised learning methods for classification and localization in histology images: a survey.
TL;DR: Results indicate that several deep learning models, and in particular WILDCAT and deep MIL can provide a high level of classification accuracy, although pixel-wise localization of cancer regions remains an issue for such images.
Posted Content
Deep Interpretable Classification and Weakly-Supervised Segmentation of Histology Images via Max-Min Uncertainty
TL;DR: High uncertainty is introduced as a criterion to localize non-discriminative regions that do not affect classifier decision, and is described with original Kullback-Leibler (KL) divergence losses evaluating the deviation of posterior predictions from the uniform distribution.
Proceedings ArticleDOI
Deep Active Learning for Joint Classification & Segmentation with Weak Annotator
TL;DR: In this article, the authors propose an active learning (AL) framework, which progressively integrates pixel-level annotations during training, and achieves state-of-the-art performance on high-resolution medical images.
Proceedings Article
Deep multi-task learning with evolving weights.
TL;DR: A multi-task learning framework that gathers weighted supervised and unsupervised tasks is used and it is proposed to evolve the weights along the learning epochs in order to avoid the break in the sequential transfer learning used in the pre-training scheme.