1 code implementation • 30 Jun 2021 • Simon Tihon, Muhammad Usama Javaid, Damien Fourure, Nicolas Posocco, Thomas Peel
Missing data is a recurrent and challenging problem, especially when using machine learning algorithms for real-world applications.
1 code implementation • 30 Jun 2021 • Damien Fourure, Muhammad Usama Javaid, Nicolas Posocco, Simon Tihon
Moreover, we show that the F1-score and the AVPR cannot be used to compare performances on different datasets as they do not reflect the intrinsic difficulty of modeling such data.
1 code implementation • 25 Jul 2017 • Damien Fourure, Rémi Emonet, Elisa Fromont, Damien Muselet, Alain Tremeau, Christian Wolf
However, for semantic image segmentation, where the task consists in providing a semantic class to each pixel of an image, feature maps reduction is harmful because it leads to a resolution loss in the output prediction.