Search

Your search keyword '"Seurin, M."' showing total 19 results
Start Over Author "Seurin, M."
19 results on '"Seurin, M."'

2. 'I'm sorry Dave, I'm afraid I can't do that' Deep Q-Learning From Forbidden Actions

Author

Seurin, M. (Mathieu), Preux, P. (Philippe), Pietquin, O. (Olivier), Université de Lille, Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 (CRIStAL), Université de Lille-Ecole Centrale de Lille-Centre National de la Recherche Scientifique (CNRS), Sequential Learning (SEQUEL), Inria Lille - Nord Europe, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 (CRIStAL), Université de Lille-Ecole Centrale de Lille-Centre National de la Recherche Scientifique (CNRS)-Université de Lille-Ecole Centrale de Lille-Centre National de la Recherche Scientifique (CNRS), Google Brain, Paris, Seurin, Mathieu, Centrale Lille-Université de Lille-Centre National de la Recherche Scientifique (CNRS), Centrale Lille-Université de Lille-Centre National de la Recherche Scientifique (CNRS)-Centrale Lille-Université de Lille-Centre National de la Recherche Scientifique (CNRS), Scool (Scool), Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL], and Sequential Learning [SEQUEL]

Subjects
[INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI], Deep Reinforcement Learning, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], [INFO.INFO-NE] Computer Science [cs]/Neural and Evolutionary Computing [cs.NE], [INFO.INFO-LG] Computer Science [cs]/Machine Learning [cs.LG], Safety, [INFO.INFO-NE]Computer Science [cs]/Neural and Evolutionary Computing [cs.NE], constraints, Q-Learning, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI]
Abstract

International audience; The use of Reinforcement Learning (RL) is still restricted to simulation or to enhance human-operated systems through recommendations. Real-world environments (e.g. industrial robots or power grids) are generally designed with safety constraints in mind implemented in the shape of valid actions masks or contingency controllers. For example, the range of motion and the angles of the motors of a robot can be limited to physical boundaries. Violating constraints thus results in rejected actions or entering in a safe mode driven by an external controller, making RL agents incapable of learning from their mistakes. In this paper, we propose a simple modification of a state-of-the-art deep RL algorithm (DQN), enabling learning from forbidden actions. To do so, the standard Q-learning update is enhanced with an extra safety loss inspired by structured classification. We empirically show that it reduces the number of hit constraints during the learning phase and accelerates convergence to near-optimal policies compared to using standard DQN. Experiments are done on a Visual Grid World Environment and Text-World domain.

Published
2019

5. Independent active decoupling circuit for RX coil on MR systems without active decoupling capabilities (0.2T) or with connecting issues (3T)

Author

Feuillet, T., Saint-Jalmes, Hervé, Seurin, M. J., Zani, M., Bordelois, A., Beuf, O., RMN et optique : De la mesure au biomarqueur, Centre de Recherche en Acquisition et Traitement de l'Image pour la Santé (CREATIS), Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Université de Lyon-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Université Jean Monnet - Saint-Étienne (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Université Jean Monnet - Saint-Étienne (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS), SARL CIRMA, CIRMA, Laboratoire Traitement du Signal et de l'Image (LTSI), Université de Rennes (UR)-Institut National de la Santé et de la Recherche Médicale (INSERM), CRLCC Eugène Marquis (CRLCC), Agilent Technologies, CBM, Universidad de Oriente, Université Jean Monnet [Saint-Étienne] (UJM)-Hospices Civils de Lyon (HCL)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Université de Lyon-Institut National des Sciences Appliquées (INSA)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Centre National de la Recherche Scientifique (CNRS)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Université Jean Monnet [Saint-Étienne] (UJM)-Hospices Civils de Lyon (HCL)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Centre National de la Recherche Scientifique (CNRS)-Institut National de la Santé et de la Recherche Médicale (INSERM), Université de Rennes 1 (UR1), and Université de Rennes (UNIV-RENNES)-Université de Rennes (UNIV-RENNES)-Institut National de la Santé et de la Recherche Médicale (INSERM)

Subjects
[SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Abstract

International audience; Some MR systems do not provide an active decoupling bias signal to RX coil and use only passive and/or geometrical decoupling. An external circuit which independently generates an active decoupling signal for RX coils for any MR system was developed. The circuit was tested on a 0.2T system with no active decoupling, and at 3T to demonstrate its interest when an additional dedicated RX coil is designed and its connection is unknown or third part coil is not allowed. Results were considered at both fields and were comparable to reference decoupling methods, in terms of delays, SNR and B1 uniformity.

Published
2014

6. Independent circuit for active decoupling of RX coil for systems without active decoupling at 0.2T or with connecting issues (3T)

Author

Feuillet, T., Seurin, M. J., Beuf, O., RMN et optique : De la mesure au biomarqueur, Centre de Recherche en Acquisition et Traitement de l'Image pour la Santé (CREATIS), Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Université de Lyon-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Hospices Civils de Lyon (HCL)-Université Jean Monnet [Saint-Étienne] (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Hospices Civils de Lyon (HCL)-Université Jean Monnet [Saint-Étienne] (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS), SARL CIRMA, and CIRMA

Subjects
[SDV.IB.IMA]Life Sciences [q-bio]/Bioengineering/Imaging
Abstract

International audience; Decoupling means are used in MRI both to protect electronics and to ensure patient safety by isolating reception (RX) coil from TX coil during transmission[1]. While widely used, some dedicated systems still do not exhibit active decoupling signal and use passive and/or geometrical configurations. Active decoupling has advantages against other methods, in terms of Signal-to-Noise Ratio (SNR) (high-Q components) or for better control of potential B1 distortions[2].An external circuit which independently generates an active decoupling signal for RX coils for any MR system was developed. This circuit was tested on a 0.2T system with no active decoupling, and at 3T to demonstrate its interest when additional RX coil is required and connection is unknown or not allowed.

Published
2013

10. Sucrose solution freezing studied by magnetic resonance imaging

Author

Mahdjoub, R., Chouvenc, P., Seurin, M. J., Andrieu, J., Briguet, A., Laboratoire Creatis, Compte Général, Centre de Recherche et d'Application en Traitement de l'Image et du Signal (CREATIS), Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Université de Lyon-Institut National des Sciences Appliquées de Lyon (INSA Lyon), and Institut National des Sciences Appliquées (INSA)-Université de Lyon-Institut National des Sciences Appliquées (INSA)-École Supérieure Chimie Physique Électronique de Lyon-Centre National de la Recherche Scientifique (CNRS)-Institut National de la Santé et de la Recherche Médicale (INSERM)

Subjects
[SDV.IB.IMA] Life Sciences [q-bio]/Bioengineering/Imaging, [SDV.IB.IMA]Life Sciences [q-bio]/Bioengineering/Imaging
Abstract

article

Published
2006

12. Coil optimization for low-field MRI: a dedicated process for small animal preclinical studies.

Author

Feuillet, T, Seurin, M-J, Leveneur, O, Viguier, E, and Beuf, O

Subjects
*MAGNETIC resonance imaging, *SIGNAL-to-noise ratio, *SALINE solutions, *TISSUE wounds, *CLINICAL trials, *LABORATORY rats
Abstract

We demonstrate a method for the fast in vivo quantification of small volumes, down to 25 µL, using low-field magnetic resonance imaging (MRI) coils. The coils were designed so as to maximize the signal-to-noise ratio (SNR) in the images. For this we developed an analytical model for describing the variations of the SNR with coil design and with size/shape suited to the object under observation. Based on the conclusions drawn from the model, the coil parameters were chosen in order to reach an SNR close to the maximum. For the validation of the model, coils were finally characterized in terms of quality factor using saline phantoms. The coil design procedure is illustrated here with two examples: first, the quantification of about 200 µL of intradermal injected gel on rabbits with a single loop surface coil and second, the imaging of the intervertebral disks in rat tails using a small volume coil to detect possible lesions. Such studies would not have been feasible for the clinical low-field MRI system at our disposal using any of the commercially available medium-sized manufactured coils. As a result of this simple optimization procedure, a wide range of applications is accessible even at low magnetic fields, leading to new opportunities for low-cost, though efficient, preclinical studies. [ABSTRACT FROM PUBLISHER]

Published
2015
Full Text
View/download PDF

15. Fours de potiers découverts à Labastide-d’Armagnac

Author

Clémens, Jacques, Jérebzoff, A., Seurin, M., Bost, Jean-Pierre, Ausonius-Institut de recherche sur l'Antiquité et le Moyen âge, Université Bordeaux Montaigne-Centre National de la Recherche Scientifique (CNRS), LabEx Sciences archéologiques de Bordeaux (LASCARBX), Université de Bordeaux (UB)-Université Bordeaux Montaigne, Université Bordeaux Montaigne-Université de Bordeaux (UB), and Université, Bordeaux Montaigne

Subjects
[SHS.HIST] Humanities and Social Sciences/History, [SHS] Humanities and Social Sciences, [SHS.HIST]Humanities and Social Sciences/History, ComputingMilieux_MISCELLANEOUS, [SHS]Humanities and Social Sciences
Abstract

International audience

Published
1977

16. Atlas des départements français d'Outre-Mer : 4. La Guyane

Author

Boyé, M., Réaud, G., Papy, G. (collab.), Seurin, M. (collab.), Lasserre, G. (ed.), Sautter, Gilles (ed.), Boyé, M. (ed.), Brasseur, Gérard (ed.), Réaud, G. (coord.), Cabaussel, G. (coord.), and Menault, J. (coord.)

Subjects
CARTE, ECONOMIE, ATLAS, HISTOIRE, ENVIRONNEMENT, POPULATION
Published
1979

19. Trabecular alveolar bone microarchitecture in the human mandible using high resolution magnetic resonance imaging.

Author

Choël L, Last D, Duboeuf F, Seurin MJ, Lissac M, Briguet A, and Guillot G

Subjects
Absorptiometry, Photon, Aged, Aged, 80 and over, Bicuspid, Cadaver, Dentition, Feasibility Studies, Female, Humans, Image Enhancement methods, Image Processing, Computer-Assisted methods, Incisor, Jaw, Edentulous pathology, Male, Middle Aged, Molar, Sex Factors, Alveolar Process ultrastructure, Bone Density physiology, Magnetic Resonance Imaging methods, Mandible ultrastructure
Abstract

Objectives: Cancellous bone microarchitecture in the mandible can influence the success of dental implant osseointegration. The aims of this study were to explore the feasibility of two-dimensional (2D) high resolution magnetic resonance imaging (MRI) for the evaluation of trabecular bone architecture and to compare architecture parameters derived from MR images between different areas in the mandible, and between sex and dental status., Methods: 45 mandibular bone specimens (8 mm thickness) were cut from 15 fresh cadavers. MR images were obtained at 2 T using a multislice 2D spin-echo sequence. After MR image binarization, histomorphometry parameters, i.e. bone area fraction (BAF), specific perimeter (La), trabecular bone width (Tb.Wi), trabecular bone separation (Tb.Sp) and trabecular network anisotropy (R), were computed from the mean intercept length. The angle between trabeculae and tooth axis and the Euler-Poincaré characteristics (EPCs) were also computed., Results: BAF, Tb.Wi and R were significantly higher in male specimens compared with female specimens, whereas Tb.Sp was significantly lower. The apparent Tb.Wi was found to be significantly higher in dentate specimens compared with edentulous ones. The highest anisotropy, corresponding to the smallest R value, was seen in the incisal specimens. The preferential orientation of the trabeculae was close to perpendicular to the tooth axis, especially in the molar specimens., Conclusion: BAF, La, Tb.Wi, Tb.Sp and R, the most variable parameters, may potentially have a relationship with the biomechanical competence of trabecular bone and play a role both in primary stabilization of dental implant and the time needed before loading.

Published
2004
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results