Zobrazeno 1 - 10
of 2 397
pro vyhledávání: '"Herbin, A."'
For classification models based on neural networks, the maximum predicted class probability is often used as a confidence score. This score rarely predicts well the probability of making a correct prediction and requires a post-processing calibration
Externí odkaz:
http://arxiv.org/abs/2411.02988
Publikováno v:
Uncertainty meets Explainability | Workshop and Tutorial @ ECML-PKDD 2023, Sep 2023, Torino, Italy
We propose to condition a generative model by a given image classifier uncertainty in order to analyze and explain its behavior. Preliminary experiments on synthetic data and a corrupted version of MNIST dataset illustrate the idea.
Externí odkaz:
http://arxiv.org/abs/2410.13871
Publikováno v:
The IJCAI-ECAI-22 Workshop on Artificial Intelligence Safety (AISafety 2022), Jul 2022, Vienna, Austria
We address in this work the question of identifying the failure conditions of a given image classifier. To do so, we exploit the capacity of producing controllable distributions of high quality image data made available by recent Generative Adversari
Externí odkaz:
http://arxiv.org/abs/2410.12814
We consider a finite volume scheme with two-point flux approximation (TPFA) to approximate a Laplace problem when the solution exhibits no more regularity than belonging to $H^1_0(\Omega)$. We establish in this case some error bounds for both the sol
Externí odkaz:
http://arxiv.org/abs/2405.16985
Publikováno v:
Generative Models for Computer Vision - CVPR 2024 Workshop, Jun 2024, Seattle, United States
Image classifiers should be used with caution in the real world. Performance evaluated on a validation set may not reflect performance in the real world. In particular, classifiers may perform well for conditions that are frequently encountered durin
Externí odkaz:
http://arxiv.org/abs/2405.02332
Publikováno v:
Atmospheric Chemistry and Physics, Vol 24, Pp 12277-12294 (2024)
Mineral dust significantly influences the Earth's climate system by affecting the radiative balance through the emission, absorption and scattering of solar and terrestrial radiation. Estimating the dust radiative effect remains challenging due to th
Externí odkaz:
https://doaj.org/article/f344a0b8a5714bbc93cb3f904b7d11a4
We consider a general linear parabolic problem with extended time boundary conditions (including initial value problems and periodic ones), and approximate it by the implicit Euler scheme in time and the Gradient Discretisation method in space; the l
Externí odkaz:
http://arxiv.org/abs/2308.10703
Autor:
Camille Herbin, Pascale Jadoul, Edouard Hosten, Amandine Gerday, Mathieu Luyckx, Jean-Luc Squifflet, Vasiliki Perlepe, Charlotte Maillard
Publikováno v:
Journal of Medical Case Reports, Vol 18, Iss 1, Pp 1-7 (2024)
Abstract Background Cervicitis, an infectious or noninfectious inflammation of the cervix, encompasses a wide range of clinical conditions, from asymptomatic infections to severe lesions, making its diagnosis difficult. Acute cervicitis may develop i
Externí odkaz:
https://doaj.org/article/e1f9a75e70d84f2ba3aecb89ff7801c3
We propose in this paper a discretization of the momentum convection operator for fluid flow simulations on quadrangular or hexahedral meshes. The space discretization is performed by the loworder nonconforming Rannacher-Turek finite element: the sca
Externí odkaz:
http://arxiv.org/abs/2301.02087
Although deep networks have shown vulnerability to evasion attacks, such attacks have usually unrealistic requirements. Recent literature discussed the possibility to remove or not some of these requirements. This paper contributes to this literature
Externí odkaz:
http://arxiv.org/abs/2212.05827