Slide
Centre Interdisciplinaire
de Recherche et d’Innovation
en Cybersécurité et Société
de Recherche et d’Innovation
en Cybersécurité et Société
1.
Hebbache, L.; Amirkhani, D.; Allili, M. S.; Hammouche, N.; Lapointe, J. -F.
Leveraging Saliency in Single-Stage Multi-Label Concrete Defect Detection Using Unmanned Aerial Vehicle Imagery Article de journal
Dans: Remote Sensing, vol. 15, no 5, 2023, ISSN: 20724292, (Publisher: MDPI).
Résumé | Liens | BibTeX | Étiquettes: Aerial vehicle, Aircraft detection, Antennas, Computational efficiency, Concrete defects, Deep learning, Defect detection, extraction, Feature extraction, Features extraction, Image acquisition, Image Enhancement, Multi-labels, One-stage concrete defect detection, Saliency, Single stage, Unmanned aerial vehicles (UAV), Unmanned areal vehicle imagery
@article{hebbache_leveraging_2023,
title = {Leveraging Saliency in Single-Stage Multi-Label Concrete Defect Detection Using Unmanned Aerial Vehicle Imagery},
author = {L. Hebbache and D. Amirkhani and M. S. Allili and N. Hammouche and J. -F. Lapointe},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85149966766&doi=10.3390%2frs15051218&partnerID=40&md5=7bf1cb3353270c696c07ff24dc24655d},
doi = {10.3390/rs15051218},
issn = {20724292},
year = {2023},
date = {2023-01-01},
journal = {Remote Sensing},
volume = {15},
number = {5},
abstract = {Visual inspection of concrete structures using Unmanned Areal Vehicle (UAV) imagery is a challenging task due to the variability of defects’ size and appearance. This paper proposes a high-performance model for automatic and fast detection of bridge concrete defects using UAV-acquired images. Our method, coined the Saliency-based Multi-label Defect Detector (SMDD-Net), combines pyramidal feature extraction and attention through a one-stage concrete defect detection model. The attention module extracts local and global saliency features, which are scaled and integrated with the pyramidal feature extraction module of the network using the max-pooling, multiplication, and residual skip connections operations. This has the effect of enhancing the localisation of small and low-contrast defects, as well as the overall accuracy of detection in varying image acquisition ranges. Finally, a multi-label loss function detection is used to identify and localise overlapping defects. The experimental results on a standard dataset and real-world images demonstrated the performance of SMDD-Net with regard to state-of-the-art techniques. The accuracy and computational efficiency of SMDD-Net make it a suitable method for UAV-based bridge structure inspection. © 2023 by the authors.},
note = {Publisher: MDPI},
keywords = {Aerial vehicle, Aircraft detection, Antennas, Computational efficiency, Concrete defects, Deep learning, Defect detection, extraction, Feature extraction, Features extraction, Image acquisition, Image Enhancement, Multi-labels, One-stage concrete defect detection, Saliency, Single stage, Unmanned aerial vehicles (UAV), Unmanned areal vehicle imagery},
pubstate = {published},
tppubtype = {article}
}
Visual inspection of concrete structures using Unmanned Areal Vehicle (UAV) imagery is a challenging task due to the variability of defects’ size and appearance. This paper proposes a high-performance model for automatic and fast detection of bridge concrete defects using UAV-acquired images. Our method, coined the Saliency-based Multi-label Defect Detector (SMDD-Net), combines pyramidal feature extraction and attention through a one-stage concrete defect detection model. The attention module extracts local and global saliency features, which are scaled and integrated with the pyramidal feature extraction module of the network using the max-pooling, multiplication, and residual skip connections operations. This has the effect of enhancing the localisation of small and low-contrast defects, as well as the overall accuracy of detection in varying image acquisition ranges. Finally, a multi-label loss function detection is used to identify and localise overlapping defects. The experimental results on a standard dataset and real-world images demonstrated the performance of SMDD-Net with regard to state-of-the-art techniques. The accuracy and computational efficiency of SMDD-Net make it a suitable method for UAV-based bridge structure inspection. © 2023 by the authors.
2.
Tremblay, L.; Chebbi, B.; Bouchard, S.
The predictive role of body image and anti-fat attitudes on attentional bias toward body area in haptic virtual reality environment Article de journal
Dans: Virtual Reality, vol. 26, no 1, p. 333–342, 2022, ISSN: 13594338, (Publisher: Springer Science and Business Media Deutschland GmbH).
Résumé | Liens | BibTeX | Étiquettes: body image, Body parts, Image Enhancement, Upper limbs, Virtual humans, virtual reality, Virtual-reality environment
@article{tremblay_predictive_2022,
title = {The predictive role of body image and anti-fat attitudes on attentional bias toward body area in haptic virtual reality environment},
author = {L. Tremblay and B. Chebbi and S. Bouchard},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85113139887&doi=10.1007%2fs10055-021-00569-4&partnerID=40&md5=3b77f665011b82e40c9ce8d5f438146a},
doi = {10.1007/s10055-021-00569-4},
issn = {13594338},
year = {2022},
date = {2022-01-01},
journal = {Virtual Reality},
volume = {26},
number = {1},
pages = {333–342},
abstract = {Evidence suggests that dissatisfaction with body image in women can be enhanced by exposure to media’s idealized images. The theory of social comparison and the avoidance hypothesis offer contradictory explanations of this relationship. We compare these two theories using a haptic virtual reality environment. We ask 42 female participants to interact with one of four types of virtual humans (VH) randomly assigned to them. The interaction task involves giving a virtual hug to a normal weight or overweight male or female VH. We verify the hypothesis that participants’ satisfaction with particular body parts and their anti-fat attitudes will determine the choice of the body area of the VH they will virtually touch. Our results show that: (1) touching VH lower torso is predicted by less anti-fat attitude, and avoidance of the upper torso and upper limb areas, and (2) touching VH shoulder and upper limbs areas is predicted by concerns with own stomach area and avoidance of VH lower torso and stomach waist areas. Our results tend to support the avoidance hypothesis as well as other research findings on anti-fat attitudes. © 2021, The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature.},
note = {Publisher: Springer Science and Business Media Deutschland GmbH},
keywords = {body image, Body parts, Image Enhancement, Upper limbs, Virtual humans, virtual reality, Virtual-reality environment},
pubstate = {published},
tppubtype = {article}
}
Evidence suggests that dissatisfaction with body image in women can be enhanced by exposure to media’s idealized images. The theory of social comparison and the avoidance hypothesis offer contradictory explanations of this relationship. We compare these two theories using a haptic virtual reality environment. We ask 42 female participants to interact with one of four types of virtual humans (VH) randomly assigned to them. The interaction task involves giving a virtual hug to a normal weight or overweight male or female VH. We verify the hypothesis that participants’ satisfaction with particular body parts and their anti-fat attitudes will determine the choice of the body area of the VH they will virtually touch. Our results show that: (1) touching VH lower torso is predicted by less anti-fat attitude, and avoidance of the upper torso and upper limb areas, and (2) touching VH shoulder and upper limbs areas is predicted by concerns with own stomach area and avoidance of VH lower torso and stomach waist areas. Our results tend to support the avoidance hypothesis as well as other research findings on anti-fat attitudes. © 2021, The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature.
3.
Allili, M. S.; Casemajor, N.; Talbi, A.
Multiple image copy detection and evolution visualisation using tree graphs Article de journal
Dans: Multimedia Tools and Applications, vol. 78, no 5, p. 6253–6275, 2019, ISSN: 13807501, (Publisher: Springer New York LLC).
Résumé | Liens | BibTeX | Étiquettes: Copy detection, Copyright protections, Copyrights, Forestry, Image copy detection, Image Enhancement, Image indexing, Image transformations, Multiple image, Original images, Tree graph, Trees (mathematics)
@article{allili_multiple_2019,
title = {Multiple image copy detection and evolution visualisation using tree graphs},
author = {M. S. Allili and N. Casemajor and A. Talbi},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85050295791&doi=10.1007%2fs11042-018-6350-5&partnerID=40&md5=21d91677491c36f971e27b7168f5b3dc},
doi = {10.1007/s11042-018-6350-5},
issn = {13807501},
year = {2019},
date = {2019-01-01},
journal = {Multimedia Tools and Applications},
volume = {78},
number = {5},
pages = {6253–6275},
abstract = {Image copy detection is an important problem for several applications such as detecting forgery to enforce copyright protection and intellectual property. One of the important problems following copy detection, however, is the assessment of the type of modifications undergone by an original image to form its copies. In this work, we propose a method for quantifying some of these modifications when multiple copies of the same image are available. We also propose an algorithm to estimate temporal precedence between images (i.e., the order of creation of the copies). Using the estimated relations, a tree graph is then built to visualize the history of evolution of the original image into its copies. Our work is important for ensuring better interpretation of image copies after their detection. It also lays a new ground for enhancing image indexing, dissemination analysis and search on the Web. © 2018, Springer Science+Business Media, LLC, part of Springer Nature.},
note = {Publisher: Springer New York LLC},
keywords = {Copy detection, Copyright protections, Copyrights, Forestry, Image copy detection, Image Enhancement, Image indexing, Image transformations, Multiple image, Original images, Tree graph, Trees (mathematics)},
pubstate = {published},
tppubtype = {article}
}
Image copy detection is an important problem for several applications such as detecting forgery to enforce copyright protection and intellectual property. One of the important problems following copy detection, however, is the assessment of the type of modifications undergone by an original image to form its copies. In this work, we propose a method for quantifying some of these modifications when multiple copies of the same image are available. We also propose an algorithm to estimate temporal precedence between images (i.e., the order of creation of the copies). Using the estimated relations, a tree graph is then built to visualize the history of evolution of the original image into its copies. Our work is important for ensuring better interpretation of image copies after their detection. It also lays a new ground for enhancing image indexing, dissemination analysis and search on the Web. © 2018, Springer Science+Business Media, LLC, part of Springer Nature.
4.
Laib, L.; Allili, M. S.; Ait-Aoudia, S.
A probabilistic topic model for event-based image classification and multi-label annotation Article de journal
Dans: Signal Processing: Image Communication, vol. 76, p. 283–294, 2019, ISSN: 09235965 (ISSN), (Publisher: Elsevier B.V.).
Résumé | Liens | BibTeX | Étiquettes: Annotation performance, Classification (of information), Convolution, Convolution neural network, Convolutional neural nets, Event classification, Event recognition, Image annotation, Image Enhancement, Latent Dirichlet allocation, Multi-label annotation, Neural networks, Probabilistic topic models, Semantics, Statistics, Topic Modeling
@article{laib_probabilistic_2019,
title = {A probabilistic topic model for event-based image classification and multi-label annotation},
author = {L. Laib and M. S. Allili and S. Ait-Aoudia},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85067936924&doi=10.1016%2fj.image.2019.05.012&partnerID=40&md5=a617885b93f3a931c6b6ce1a165f940b},
doi = {10.1016/j.image.2019.05.012},
issn = {09235965 (ISSN)},
year = {2019},
date = {2019-01-01},
journal = {Signal Processing: Image Communication},
volume = {76},
pages = {283–294},
abstract = {We propose an enhanced latent topic model based on latent Dirichlet allocation and convolutional neural nets for event classification and annotation in images. Our model builds on the semantic structure relating events, objects and scenes in images. Based on initial labels extracted from convolution neural networks (CNNs), and possibly user-defined tags, we estimate the event category and final annotation of an image through a refinement process based on the expectation–maximization (EM)algorithm. The EM steps allow to progressively ascertain the class category and refine the final annotation of the image. Our model can be thought of as a two-level annotation system, where the first level derives the image event from CNN labels and image tags and the second level derives the final annotation consisting of event-related objects/scenes. Experimental results show that the proposed model yields better classification and annotation performance in the two standard datasets: UIUC-Sports and WIDER. © 2019 Elsevier B.V.},
note = {Publisher: Elsevier B.V.},
keywords = {Annotation performance, Classification (of information), Convolution, Convolution neural network, Convolutional neural nets, Event classification, Event recognition, Image annotation, Image Enhancement, Latent Dirichlet allocation, Multi-label annotation, Neural networks, Probabilistic topic models, Semantics, Statistics, Topic Modeling},
pubstate = {published},
tppubtype = {article}
}
We propose an enhanced latent topic model based on latent Dirichlet allocation and convolutional neural nets for event classification and annotation in images. Our model builds on the semantic structure relating events, objects and scenes in images. Based on initial labels extracted from convolution neural networks (CNNs), and possibly user-defined tags, we estimate the event category and final annotation of an image through a refinement process based on the expectation–maximization (EM)algorithm. The EM steps allow to progressively ascertain the class category and refine the final annotation of the image. Our model can be thought of as a two-level annotation system, where the first level derives the image event from CNN labels and image tags and the second level derives the final annotation consisting of event-related objects/scenes. Experimental results show that the proposed model yields better classification and annotation performance in the two standard datasets: UIUC-Sports and WIDER. © 2019 Elsevier B.V.
5.
Allili, M. S.; Ziou, D.
Likelihood-based feature relevance for figure-ground segmentation in images and videos Article de journal
Dans: Neurocomputing, vol. 167, p. 658–670, 2015, ISSN: 09252312, (Publisher: Elsevier).
Résumé | Liens | BibTeX | Étiquettes: accuracy, algorithm, article, calculation, Feature relevance, Figure-ground segmentations, Gaussian mixture model (GMMs), Image analysis, Image Enhancement, image quality, Image segmentation, Level Set, linear system, mathematical analysis, mathematical model, Negative examples, priority journal, Video cameras, videorecording
@article{allili_likelihood-based_2015,
title = {Likelihood-based feature relevance for figure-ground segmentation in images and videos},
author = {M. S. Allili and D. Ziou},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84952631642&doi=10.1016%2fj.neucom.2015.04.015&partnerID=40&md5=833948d0784e0dc42c2245b9343971dd},
doi = {10.1016/j.neucom.2015.04.015},
issn = {09252312},
year = {2015},
date = {2015-01-01},
journal = {Neurocomputing},
volume = {167},
pages = {658–670},
abstract = {We propose an efficient method for image/video figure-ground segmentation using feature relevance (FR) and active contours. Given a set of positive and negative examples of a specific foreground (an object of interest (OOI) in an image or a tracked objet in a video), we first learn the foreground distribution model and its characteristic features that best discriminate it from its contextual background. For this goal, an objective function based on feature likelihood ratio is proposed for supervised FR computation. FR is then incorporated in foreground segmentation of new images and videos using level sets and energy minimization. We show the effectiveness of our approach on several examples of image/video figure-ground segmentation. © 2015 Elsevier B.V.},
note = {Publisher: Elsevier},
keywords = {accuracy, algorithm, article, calculation, Feature relevance, Figure-ground segmentations, Gaussian mixture model (GMMs), Image analysis, Image Enhancement, image quality, Image segmentation, Level Set, linear system, mathematical analysis, mathematical model, Negative examples, priority journal, Video cameras, videorecording},
pubstate = {published},
tppubtype = {article}
}
We propose an efficient method for image/video figure-ground segmentation using feature relevance (FR) and active contours. Given a set of positive and negative examples of a specific foreground (an object of interest (OOI) in an image or a tracked objet in a video), we first learn the foreground distribution model and its characteristic features that best discriminate it from its contextual background. For this goal, an objective function based on feature likelihood ratio is proposed for supervised FR computation. FR is then incorporated in foreground segmentation of new images and videos using level sets and energy minimization. We show the effectiveness of our approach on several examples of image/video figure-ground segmentation. © 2015 Elsevier B.V.
6.
Allili, M. S.
Wavelet modeling using finite mixtures of generalized Gaussian distributions: Application to texture discrimination and retrieval Article de journal
Dans: IEEE Transactions on Image Processing, vol. 21, no 4, p. 1452–1464, 2012, ISSN: 10577149.
Résumé | Liens | BibTeX | Étiquettes: algorithm, Algorithms, article, Automated, automated pattern recognition, computer assisted diagnosis, Computer Simulation, Computer-Assisted, Data Interpretation, Finite mixtures, Generalized Gaussian, Generalized Gaussian Distributions, Image Enhancement, Image Interpretation, Image segmentation, Imaging, Kullback Leibler divergence, Marginal distribution, methodology, Mixtures, Models, Monte Carlo methods, Monte Carlo sampling, Normal Distribution, Pattern Recognition, Performance improvements, reproducibility, Reproducibility of Results, Sensitivity and Specificity, Similarity measure, State-of-the-art approach, Statistical, statistical analysis, statistical model, Texture data set, Texture discrimination, Texture modeling, Textures, three dimensional imaging, Three-Dimensional, Wavelet Analysis, Wavelet coefficients, Wavelet decomposition, Wavelet modeling
@article{allili_wavelet_2012,
title = {Wavelet modeling using finite mixtures of generalized Gaussian distributions: Application to texture discrimination and retrieval},
author = {M. S. Allili},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84859096106&doi=10.1109%2fTIP.2011.2170701&partnerID=40&md5=0420facdc04978ad84bea3126bc1183a},
doi = {10.1109/TIP.2011.2170701},
issn = {10577149},
year = {2012},
date = {2012-01-01},
journal = {IEEE Transactions on Image Processing},
volume = {21},
number = {4},
pages = {1452–1464},
abstract = {This paper addresses statistical-based texture modeling using wavelets. We propose a new approach to represent the marginal distribution of the wavelet coefficients using finite mixtures of generalized Gaussian (MoGG) distributions. The MoGG captures a wide range of histogram shapes, which provides better description and discrimination of texture than using single probability density functions (pdf's), as proposed by recent state-of-the-art approaches. Moreover, we propose a model similarity measure based on Kullback-Leibler divergence (KLD) approximation using Monte Carlo sampling methods. Through experiments on two popular texture data sets, we show that our approach yields significant performance improvements for texture discrimination and retrieval, as compared with recent methods of statistical-based wavelet modeling. © 2011 IEEE.},
keywords = {algorithm, Algorithms, article, Automated, automated pattern recognition, computer assisted diagnosis, Computer Simulation, Computer-Assisted, Data Interpretation, Finite mixtures, Generalized Gaussian, Generalized Gaussian Distributions, Image Enhancement, Image Interpretation, Image segmentation, Imaging, Kullback Leibler divergence, Marginal distribution, methodology, Mixtures, Models, Monte Carlo methods, Monte Carlo sampling, Normal Distribution, Pattern Recognition, Performance improvements, reproducibility, Reproducibility of Results, Sensitivity and Specificity, Similarity measure, State-of-the-art approach, Statistical, statistical analysis, statistical model, Texture data set, Texture discrimination, Texture modeling, Textures, three dimensional imaging, Three-Dimensional, Wavelet Analysis, Wavelet coefficients, Wavelet decomposition, Wavelet modeling},
pubstate = {published},
tppubtype = {article}
}
This paper addresses statistical-based texture modeling using wavelets. We propose a new approach to represent the marginal distribution of the wavelet coefficients using finite mixtures of generalized Gaussian (MoGG) distributions. The MoGG captures a wide range of histogram shapes, which provides better description and discrimination of texture than using single probability density functions (pdf's), as proposed by recent state-of-the-art approaches. Moreover, we propose a model similarity measure based on Kullback-Leibler divergence (KLD) approximation using Monte Carlo sampling methods. Through experiments on two popular texture data sets, we show that our approach yields significant performance improvements for texture discrimination and retrieval, as compared with recent methods of statistical-based wavelet modeling. © 2011 IEEE.