Search

Your search keyword '"Rémi Ratajczak"' showing total 4 results
Start Over Author "Rémi Ratajczak" Language undetermined
4 results on '"Rémi Ratajczak"'

1. Semantic Segmentation Post-processing with Colorized Pairwise Potentials and Deep Edges

Author

Laure Tougne, Elodie Faure, Rémi Ratajczak, Carlos Crispim, Béatrice Fervers, Extraction de Caractéristiques et Identification (imagine), Laboratoire d'InfoRmatique en Image et Systèmes d'information (LIRIS), 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)-Centre National de la Recherche Scientifique (CNRS)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-École Centrale de Lyon (ECL), Université de Lyon-Université Lumière - Lyon 2 (UL2)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Université Lumière - Lyon 2 (UL2), Département cancer environnement (Centre Léon Bérard - Lyon), Centre Léon Bérard [Lyon], Agence de l'Environnement et de la Maîtrise de l'Energie (ADEME), LABEX IMU (ANR-10-LABX-0088/ ANR-11-IDEX-0007)., ADEME (N TEZ17-42), Centre Léon Bérard, IEEE, TESTIS, and IMU GOURAMIC

Subjects
Conditional random field, Computer science, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, 0211 other engineering and technologies, 02 engineering and technology, Grayscale, Deep Edges, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], Segmentation, Deep Learning, Land Use, 0202 electrical engineering, electronic engineering, information engineering, 021101 geological & geomatics engineering, Pixel, business.industry, Deep learning, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], Pattern recognition, Object (computer science), [SDE.ES]Environmental Sciences/Environmental and Society, Panchromatic film, [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], 020201 artificial intelligence & image processing, Pairwise comparison, Artificial intelligence, business
Abstract

International audience; Semantic segmentation is the task of assigning a label to each pixel in an image, providing high level insights to a wide range of end-user applications like autonomous driving, medical imaging and land use mapping. However, semantic segmentation results are not always consistent with the object boundaries and may sometimes lack of spatial consistency. To solve these problems, post-processing algorithms have been proposed, paving the way for more robust pipelines. In this work, we study a novel post-processing approach to enhance semantic segmentation of panchromatic aerial images based on unsupervised colorization and deep edge superpixels. In particular, we propose to assess whether applying a colorization algorithm could enhance the strength of the pairwise potentials used in an extended dense conditional random field. We present experiments on recent aerial color images that we convert to grayscale before colorization, allowing us to assess how colorized representations impact post-processing when compared to real color and panchromatic representations.

Published
2020

2. Pseudo-Cyclic Network for Unsupervised Colorization with Handcrafted Translation and Output Spatial Pyramids

Author

Rémi Ratajczak, Laure Tougne, Béatrice Fervers, Carlos Crispim-Junior, Elodie Faure, Agence de l'Environnement et de la Maîtrise de l'Energie (ADEME), Extraction de Caractéristiques et Identification (imagine), Laboratoire d'InfoRmatique en Image et Systèmes d'information (LIRIS), 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)-Centre National de la Recherche Scientifique (CNRS)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-École Centrale de Lyon (ECL), Université de Lyon-Université Lumière - Lyon 2 (UL2)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Université Lumière - Lyon 2 (UL2), Département cancer environnement (Centre Léon Bérard - Lyon), and Centre Léon Bérard [Lyon]

Subjects
Artificial neural network, Computer science, business.industry, [INFO.INFO-MM]Computer Science [cs]/Multimedia [cs.MM], ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, Pattern recognition, 02 engineering and technology, [INFO.INFO-NE]Computer Science [cs]/Neural and Evolutionary Computing [cs.NE], 010501 environmental sciences, Color space, Translation (geometry), 01 natural sciences, Grayscale, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], 0202 electrical engineering, electronic engineering, information engineering, 020201 artificial intelligence & image processing, State (computer science), Artificial intelligence, business, 0105 earth and related environmental sciences, Generator (mathematics), Complement (set theory)
Abstract

International audience; We present a novel pseudo-cyclic adversarial learning approach for unsupervised colorization of grayscale images. We investigate the use of a non-trainable, lightweight and well-defined Handcrafted Translation to enforce the generation of realistic images and replace one of the two deep convolutional generative adversarial neural networks classically used in cyclic models. Additionally, we propose to use Output Spatial Pyramids to jointly constrain the deep latent spaces of an encoder-decoder generator to preserve spatial structures and improve the quality of the generated images. We demonstrate the interest of our approach compared with the state of the art on standard datasets (paintings, landscapes, aerial, thumbnails) that we modified for the purpose of colorization. We evaluate colorization quality of the generated images along the training with deterministic and reproducible criteria. In complement, we demonstrate the ability of our method to generate representations that are prone to make a classification network generalize well to slightly different color spaces. We believe our approach has potential applications in arts and cultural heritage to produce alternative representations without requiring paired data.

Published
2019

3. Toward An Unsupervised Colorization Framework for Historical Land Use Classification

Author

Béatrice Fervers, Laure Tougne, Rémi Ratajczak, Carlos Crispim-Junior, Elodie Faure, Extraction de Caractéristiques et Identification (imagine), Laboratoire d'InfoRmatique en Image et Systèmes d'information (LIRIS), 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)-Centre National de la Recherche Scientifique (CNRS)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-École Centrale de Lyon (ECL), Université de Lyon-Université Lumière - Lyon 2 (UL2)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Université Lumière - Lyon 2 (UL2), Département cancer environnement (Centre Léon Bérard - Lyon), Centre Léon Bérard [Lyon], Agence de l'Environnement et de la Maîtrise de l'Energie (ADEME), LABEX IMU (ANR-10-LABX-0088/ ANR-11-IDEX-0007)., ADEME (N TEZ17-42), Centre Léon Bérard, IEEE, TESTIS, and IMU GOURAMIC

Subjects
Computer science, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, 02 engineering and technology, Texture (geology), [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], Deep Learning, Land Use, 0202 electrical engineering, electronic engineering, information engineering, Colorization, Land use, Artificial neural network, Texture filters, business.industry, Deep learning, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], Pattern recognition, 15. Life on land, Index Terms-Colorization, Classification, [SDE.ES]Environmental Sciences/Environmental and Society, GAN, Visualization, Historical Land Use, [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], 020201 artificial intelligence & image processing, Artificial intelligence, business
Abstract

International audience; We present an unsupervised colorization framework to improve both the visualization and the automatic land use classification of historical aerial images. We introduce a novel algorithm built upon a cyclic generative adversarial neural network and a texture replacement method to homogeneously and automatically colorize unpaired VHR images. We apply our framework on historical aerial images acquired in France between 1970 and 1990. We demonstrate that our approach helps to disentangle hard to classify land use classes and hence improves the overall land use classification.

Published
2019

4. A Fast Audiovisual Attention Model for Human Detection and Localization on a Companion Robot

Author

Rémi Ratajczak, denis pellerin, Quentin Labourey, Catherine Garbay, GIPSA - Architecture, Géométrie, Perception, Images, Gestes (GIPSA-AGPIG), Département Images et Signal (GIPSA-DIS), Grenoble Images Parole Signal Automatique (GIPSA-lab ), Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes [2016-2019] (UGA [2016-2019])-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes [2016-2019] (UGA [2016-2019])-Grenoble Images Parole Signal Automatique (GIPSA-lab ), Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes [2016-2019] (UGA [2016-2019])-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes [2016-2019] (UGA [2016-2019]), Analyse de données, Modélisation et Apprentissage automatique [Grenoble] (AMA ), Laboratoire d'Informatique de Grenoble (LIG ), Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes [2016-2019] (UGA [2016-2019])-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes [2016-2019] (UGA [2016-2019]), Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes [2016-2019] (UGA [2016-2019]), IARIA, ATTENTIVE, and ANR-11-LABX-0025,PERSYVAL-lab,Systemes et Algorithmes Pervasifs au confluent des mondes physique et numérique(2011)

Subjects
audiovisual attention, human localization, saliency, RGB-D, companion robot, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Abstract

International audience; This paper describes a fast audiovisual attention model applied to human detection and localization on a companion robot. Its originality lies in combining static and dynamic modalities over two analysis paths in order to guide the robot's gaze towards the most probable human beings' locations based on the concept of saliency. Visual, depth and audio data are acquired using a RGB-D camera and two horizontal microphones. Adapted state-of-the-art methods are used to extract relevant information and fuse them together via two dimensional gaussian representations. The obtained saliency map represents human positions as the most salient areas. Experiments have shown that the proposed model can provide a mean F-measure of 66 percent with a mean precision of 77 percent for human localization using bounding box areas on 10 manually annotated videos. The corresponding algorithm is able to process 70 frames per second on the robot.

Catalog

Books, media, physical & digital resources
See catalog results