Tracking biodiversity shifts is central to understanding past, present, and future global changes... more Tracking biodiversity shifts is central to understanding past, present, and future global changes. Recent advances in bioacoustics and the low cost of high‐quality automatic recorders are revolutionizing studies in biogeography and community and behavioral ecology with a robust assessment of phenology, species occurrence, and individual activity. This large volume of acoustic recordings has recently generated a plethora of datasets that can now be handled automatically, mostly via big data methods such as deep learning. These approaches need high‐quality annotations to classify and detect recorded sounds efficiently. However, very few strongly annotated datasets—that is, with detailed information on start and end time of each vocalization—are openly accessible to the public. Moreover, these datasets mostly cover temperate species and are usually limited to a single year of recordings. Here, we present ArcticBirdSounds, the first open‐access, multisite, and multiyear strongly annotated dataset of arctic bird vocalizations. ArcticBirdSounds offers 20 h of annotated recordings over 2 years (2018, 2019), taken from 15 distinct plots within six locations across the Arctic, from Alaska to Greenland. Recordings cover the arctic vertebrates' breeding period and are evenly spaced during the day; they capture most species breeding there with 12,933 temporal annotations in 49 classes of sounds. While these data can be used for many pressing ecological questions, it is also a unique resource for methodological development to help meet the challenges of fast ecosystem transformations such as those happening in the Arctic. All data, including audio files, annotation files, and companion spreadsheets, are available in an Open Science Framework repository published under a CC BY 4.0 License.
bioRxiv (Cold Spring Harbor Laboratory), May 30, 2018
A lot of hype has recently been generated around deep learning, a group of artificial intelligenc... more A lot of hype has recently been generated around deep learning, a group of artificial intelligence approaches able to break accuracy records in pattern recognition. Over the course of just a few years, deep learning revolutionized several research fields such as bioinformatics or medicine. Yet such a surge of tools and knowledge is still in its infancy in ecology despite the ever-growing size and the complexity of ecological datasets. Here we performed a literature review of deep learning implementations in ecology to identify its benefits in most ecological disciplines, even in applied ecology, up to decision makers and conservationists alike. We also provide guidelines on useful resources and recommendations for ecologists to start adding deep learning to their toolkit. At a time when automatic monitoring of populations and ecosystems generates a vast amount of data that cannot be processed by humans anymore, deep learning could become a necessity in ecology.
Background: Many apps have been designed to remotely assess clinical status and monitor symptom e... more Background: Many apps have been designed to remotely assess clinical status and monitor symptom evolution in persons with multiple sclerosis (MS). These may one day serve as an adjunct for in-person assessment of persons with MS, providing valuable insight into the disease course that is not well captured by cross-sectional snapshots obtained from clinic visits. Objective: This study sought to review the current literature surrounding apps used for remote monitoring of persons with MS. Methods: A scoping review of published articles was conducted to identify and evaluate the literature published regarding the use of apps for monitoring of persons with MS. PubMed/Medline, EMBASE, CINAHL, and Cochrane databases were searched from inception to January 2022. Cohort studies, feasibility studies, and randomized controlled trials were included in this review. All pediatric studies, single case studies, poster presentations, opinion pieces, and commentaries were excluded. Studies were assessed for risk of bias using the Scottish Intercollegiate Guidelines Network, when applicable. Key findings were grouped in categories (convergence to neurological exam, feasibility of implementation, impact of weather, and practice effect), and trends are presented. In a parallel systematic search, the Canadian Apple App Store and Google Play Store were searched to identify relevant apps that are available but have yet to be formally studied and published in peer-reviewed publications. Results: We included 18 articles and 18 apps. Although many MS-related apps exist, only 10 apps had published literature supporting their use. Convergence between app-based testing and the neurological exam was examined in 12 articles. Most app-based tests focused on physical disability and cognition, although other domains such as ambulation, balance, visual acuity, and fatigue were also evaluated. Overall, correlations between the app versions of standardized tests and their traditional counterparts were moderate to strong. Some novel app-based tests had a stronger correlation with clinician-derived outcomes than traditional testing. App-based testing correlated well with the Multiple Sclerosis Functional Composite but less so with the Expanded Disability Status Scale; the latter correlated to a greater extent with patient quality of life questionnaire scores. Conclusions: Although limited by a small number of included studies and study heterogeneity, the findings of this study suggest that app-based testing demonstrates adequate convergence to traditional in-person assessment and may be used as an adjunct to and perhaps in lieu of specific neurological exam metrics documented at clinic visits, particularly if the latter is not readily accessible for persons with MS.
BACKGROUND Many apps designed to remotely assess clinical status and monitor symptom evolution in... more BACKGROUND Many apps designed to remotely assess clinical status and monitor symptom evolution in persons with multiple sclerosis (MS) exist. These may one day serve as an adjunct for in-person testing of persons with MS, providing valuable insight into disease course that is not accounted for by snapshots obtained from clinic visits. OBJECTIVE This study sought to review the current literature surrounding apps used for remote monitoring of persons with MS. METHODS A systematic review of published articles was conducted in January 2022 in 4 databases. The Apple App Store and the Google Play Store were also searched, and relevant apps were identified. RESULTS Eighteen articles and eighteen apps were included in this study. Although many MS-related apps exist, only ten apps were identified in the systematic review to have published literature surrounding their use. Twelve articles looked at convergence between app-based testing and neurological exam. Overall, correlation between app v...
2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017
The vigor of a tree defines its ability to grow and is associated with its productivity ([1]; [2]... more The vigor of a tree defines its ability to grow and is associated with its productivity ([1]; [2]). The measurement of a tree's vigor makes it possible to evaluate its general state of health and its development over time. This task is traditionally performed by identifying defects affecting the vigor such as visible signs and symptoms on the tree, like fungi on the main stem, significant forks, cracks, wounds, competition and the percentage of living crown ([2]). The vigor and quality of trees are important factors used by forest managers in the application of silvicultural prescriptions at stand and landscape levels. The estimation of the level of risk of vigor loss in forest stands thus becomes a major and essential issue for the management and the improvement of the productivity of forests in a context of sustainable development.
Methods in Ecology and Evolution This article has been accepted for publication and undergone ful... more Methods in Ecology and Evolution This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as
This paper presents a new steganalysis method that uses a deep belief network (DBN) as a classifi... more This paper presents a new steganalysis method that uses a deep belief network (DBN) as a classifier for audio files. It has been tested on three steganographic techniques: StegHide, Hide4PGP and FreqSteg. The results were compared to two other existing robust steganalysis methods based on support vector machines (SVMs) and Gaussian mixture models (GMMs). Afterwards, another classification task aiming at identifying the type of steganographic applied or not to the speech signal was carried out. The results of this four-way classification show that in most cases, the proposed DBN-based steganalysis method gives higher classification rates than the two other steganalysis methods based on SVMs and GMMs.
IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium, 2008
ABSTRACT The present study develops a method for aboveground forest biomass mapping from Ikonos i... more ABSTRACT The present study develops a method for aboveground forest biomass mapping from Ikonos imagery and geospatial data. Reference biomass values by group of species were estimated using Ker's equations and inventory data from permanent sample plots (PEP) of 400 m2. A supervised classification of the Ikonos image, based on the maximum likelihood method presenting the five species groups inventoried in the field study, was carried out. Thereafter, various vegetation indices and texture parameters were extracted from the Ikonos image. The extracted Ikonos data were then combined with geospatial data at the same 1 m spatial resolution. Inventory plots biomass values estimated by group of species were used for the neural networks model (Multi-layer Perceptron) training with the backpropagation algorithm. Thereafter, biomass values for sample pixels generated randomly by group of species were predicted with the Multi-layer Perceptron. Then, sample pixels biomass values of each group were used to derive biomass values of other pixels of the same species group by interpolation with the ordinary kriging method using five different variogram models. The Gaussian variogram model yielded the best biomass estimates by comparison with reference biomass values, with percentages of residual errors ranging between 2,6 and 9,8% (absolute value) and percentages of RMSE (root mean square error) ranging between 17.2 and 61.1%.
This paper describes a method to compute dense correspondences (disparity map) between two images... more This paper describes a method to compute dense correspondences (disparity map) between two images in a stereo pair in case of significant displacements. In order to deal with large amplitude disparity fields, a multi-resolution technique called block match ing have been used in the past(1,2). Although quite successful, this technique cann ot easily cope with disparity discontinuities as it does
2005 ICSC Congress on Computational Intelligence Methods and Applications
A template-based technique for automatic recognition of birdsong syllables is presented. This tec... more A template-based technique for automatic recognition of birdsong syllables is presented. This technique combines time delay neural networks (TDNNs) with an autoregressive (AR) version of the backpropagation algorithm in order to improve the accuracy of bird species identification. The proposed neural network structure (AR-TDNN) has the advantage of dealing with a pattern classification of syllable alphabet and also of capturing
2012 11th International Conference on Information Science, Signal Processing and their Applications (ISSPA), 2012
Content-based indexing of images consists in extracting visual information from digital images (s... more Content-based indexing of images consists in extracting visual information from digital images (such as pixels, colors, objects, shapes, etc.), and can be performed automatically and fast by computers. This work focuses on color indexing of images. A statistical algorithm called Mean-Shift is used to modelize the color distributions of images as two-dimensional Gaussian kernels. The experiments have been performed on the COIL1 database and the results show that the proposed method compares well to other content-based retrieval methods.
2011 Seventh International Conference on Signal Image Technology & Internet-Based Systems, 2011
The topic of this paper is Content-Based Image Retrieval (CBIR) based on colors as a content imag... more The topic of this paper is Content-Based Image Retrieval (CBIR) based on colors as a content image descriptor. The tool we developed to that purpose modelizes the colors of an image as a set of 2D Gaussian distributions based on weighted color histograms. Then, given a reference image proposed by a user, the system can automatically classify the image and
2012 IEEE International Geoscience and Remote Sensing Symposium, 2012
This study proposes an approach which simultaneously uses spatial information and polarimetric da... more This study proposes an approach which simultaneously uses spatial information and polarimetric data from a RADARSAT-2 quad-polarization satellite image for forest tree species classification. The study area is near the Gounamitz River located in northwestern New Brunswick (Canada). After geometric correction of the image, two statistical models were used for the classification: (1) a Markov random fields model based on an initial segmentation provided by the K-means algorithm to account for the spatial statistical dependencies between adjacent sites; and (2) a K-distribution model with, as parameters, the covariance matrix containing all of the polarimetric information. The classification was optimized using the stochastic simulated annealing algorithm. Validation of the results was performed by comparison with field inventory measurements. Variation of the backscattering coefficient c° obtained for the RADARSAT-2 quad-polarization SAR image with incidence angles of 26 0 and 45 ° ranged from 1 and 3 dB for the different tree species. The results of average and overall accuracies of the classification were respectively 77.13% and 72.35% for the 26° incidence angle image compared to 81.47% and 79.12% for the 45°incidence angle.
2007 IEEE International Geoscience and Remote Sensing Symposium, 2007
ABSTRACT The present study develops an approach for the estimation of aboveground forest biomass ... more ABSTRACT The present study develops an approach for the estimation of aboveground forest biomass based on neural networks, using Ikonos satellite image data and multi-source geo-scientific data. Two methods of aboveground forest biomass estimation were compared: multiple regressions and the neural networks. Percentages of residual errors of the neural networks biomass estimates were lower than 1% for all groups of species, except for "intolerant hardwood" which had a percentage of 3.2% for the 9-18 cm DBH class. Percentages of residual errors of biomass estimates were higher with the quadratic multiple regression approach than with the neural networks, particularly for "intolerant hardwood" where a value of 51.41% was observed for the 19-40+ cm DBH class. Root mean square error values (RMSE) calculated from biomass estimates resulting from the neural networks approach were lower than those computed with estimates of the quadratic multiple regressions model, for all groups of species.
2016 IEEE Congress on Evolutionary Computation (CEC), 2016
This paper presents a new method to train Restricted Boltzmann Machines (RBMs) using Evolutionary... more This paper presents a new method to train Restricted Boltzmann Machines (RBMs) using Evolutionary Algorithms (EAs), where RBMs are used in the first step of a steganalysis tool for speech/audio files. The following EAs have been tested: Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Artificial Bees Colony (ABC) and Cat Swarm Optimization (CSO). Our method has been tested with three steganographic techniques: StegHide, Hide4PGP, and FreqSteg. A fourth technique combining the three steganographic methods has also been tested. The results are compared to the conventional contrastive divergence learning algorithm. All EAs outperform the contrastive divergence algorithm.
A lot of hype has recently been generated around deep learning, a group of artificial intelligenc... more A lot of hype has recently been generated around deep learning, a group of artificial intelligence approaches able to break accuracy records in pattern recognition. Over the course of just a few years, deep learning revolutionized several research fields such as bioinformatics or medicine. Yet such a surge of tools and knowledge is still in its infancy in ecology despite the ever-growing size and the complexity of ecological datasets. Here we performed a literature review of deep learning implementations in ecology to identify its benefits in most ecological disciplines, even in applied ecology, up to decision makers and conservationists alike. We also provide guidelines on useful resources and recommendations for ecologists to start adding deep learning to their toolkit. At a time when automatic monitoring of populations and ecosystems generates a vast amount of data that cannot be processed by humans anymore, deep learning could become a necessity in ecology.
Le radar a synthese d'ouverture (RSO) est un puissant outil de teledetection capable de fourn... more Le radar a synthese d'ouverture (RSO) est un puissant outil de teledetection capable de fournir des images a haute resolution de la surface terrestre par tout temps, de jour comme de nuit. Les images radar sont caracterisees par une forte granulation appelee speckle, inherente a la formation meme des donnees radar et qui rend extremement difficile leur interpretation automatique. L'objectif de cette these concerne la radargrammetrie ou stereovision radar, qui consiste a produire des modeles d'elevation de terrain a partir de paires d'images RSO. Pour cela il faut passer notamment par une phase-cle appelee appariement qui peut se faire a court terme en deux etapes : on effectue d'abord un filtrage du speckle, c'est-a-dire une estimation des reflectivites radar sous-jacentes de la surface imagee. Ceci permet alors d'appliquer aux images filtrees des methodes de traitement d'images eprouvees en imagerie optique. Cette approche en deux temps est cependant...
Tracking biodiversity shifts is central to understanding past, present, and future global changes... more Tracking biodiversity shifts is central to understanding past, present, and future global changes. Recent advances in bioacoustics and the low cost of high‐quality automatic recorders are revolutionizing studies in biogeography and community and behavioral ecology with a robust assessment of phenology, species occurrence, and individual activity. This large volume of acoustic recordings has recently generated a plethora of datasets that can now be handled automatically, mostly via big data methods such as deep learning. These approaches need high‐quality annotations to classify and detect recorded sounds efficiently. However, very few strongly annotated datasets—that is, with detailed information on start and end time of each vocalization—are openly accessible to the public. Moreover, these datasets mostly cover temperate species and are usually limited to a single year of recordings. Here, we present ArcticBirdSounds, the first open‐access, multisite, and multiyear strongly annotated dataset of arctic bird vocalizations. ArcticBirdSounds offers 20 h of annotated recordings over 2 years (2018, 2019), taken from 15 distinct plots within six locations across the Arctic, from Alaska to Greenland. Recordings cover the arctic vertebrates' breeding period and are evenly spaced during the day; they capture most species breeding there with 12,933 temporal annotations in 49 classes of sounds. While these data can be used for many pressing ecological questions, it is also a unique resource for methodological development to help meet the challenges of fast ecosystem transformations such as those happening in the Arctic. All data, including audio files, annotation files, and companion spreadsheets, are available in an Open Science Framework repository published under a CC BY 4.0 License.
bioRxiv (Cold Spring Harbor Laboratory), May 30, 2018
A lot of hype has recently been generated around deep learning, a group of artificial intelligenc... more A lot of hype has recently been generated around deep learning, a group of artificial intelligence approaches able to break accuracy records in pattern recognition. Over the course of just a few years, deep learning revolutionized several research fields such as bioinformatics or medicine. Yet such a surge of tools and knowledge is still in its infancy in ecology despite the ever-growing size and the complexity of ecological datasets. Here we performed a literature review of deep learning implementations in ecology to identify its benefits in most ecological disciplines, even in applied ecology, up to decision makers and conservationists alike. We also provide guidelines on useful resources and recommendations for ecologists to start adding deep learning to their toolkit. At a time when automatic monitoring of populations and ecosystems generates a vast amount of data that cannot be processed by humans anymore, deep learning could become a necessity in ecology.
Background: Many apps have been designed to remotely assess clinical status and monitor symptom e... more Background: Many apps have been designed to remotely assess clinical status and monitor symptom evolution in persons with multiple sclerosis (MS). These may one day serve as an adjunct for in-person assessment of persons with MS, providing valuable insight into the disease course that is not well captured by cross-sectional snapshots obtained from clinic visits. Objective: This study sought to review the current literature surrounding apps used for remote monitoring of persons with MS. Methods: A scoping review of published articles was conducted to identify and evaluate the literature published regarding the use of apps for monitoring of persons with MS. PubMed/Medline, EMBASE, CINAHL, and Cochrane databases were searched from inception to January 2022. Cohort studies, feasibility studies, and randomized controlled trials were included in this review. All pediatric studies, single case studies, poster presentations, opinion pieces, and commentaries were excluded. Studies were assessed for risk of bias using the Scottish Intercollegiate Guidelines Network, when applicable. Key findings were grouped in categories (convergence to neurological exam, feasibility of implementation, impact of weather, and practice effect), and trends are presented. In a parallel systematic search, the Canadian Apple App Store and Google Play Store were searched to identify relevant apps that are available but have yet to be formally studied and published in peer-reviewed publications. Results: We included 18 articles and 18 apps. Although many MS-related apps exist, only 10 apps had published literature supporting their use. Convergence between app-based testing and the neurological exam was examined in 12 articles. Most app-based tests focused on physical disability and cognition, although other domains such as ambulation, balance, visual acuity, and fatigue were also evaluated. Overall, correlations between the app versions of standardized tests and their traditional counterparts were moderate to strong. Some novel app-based tests had a stronger correlation with clinician-derived outcomes than traditional testing. App-based testing correlated well with the Multiple Sclerosis Functional Composite but less so with the Expanded Disability Status Scale; the latter correlated to a greater extent with patient quality of life questionnaire scores. Conclusions: Although limited by a small number of included studies and study heterogeneity, the findings of this study suggest that app-based testing demonstrates adequate convergence to traditional in-person assessment and may be used as an adjunct to and perhaps in lieu of specific neurological exam metrics documented at clinic visits, particularly if the latter is not readily accessible for persons with MS.
BACKGROUND Many apps designed to remotely assess clinical status and monitor symptom evolution in... more BACKGROUND Many apps designed to remotely assess clinical status and monitor symptom evolution in persons with multiple sclerosis (MS) exist. These may one day serve as an adjunct for in-person testing of persons with MS, providing valuable insight into disease course that is not accounted for by snapshots obtained from clinic visits. OBJECTIVE This study sought to review the current literature surrounding apps used for remote monitoring of persons with MS. METHODS A systematic review of published articles was conducted in January 2022 in 4 databases. The Apple App Store and the Google Play Store were also searched, and relevant apps were identified. RESULTS Eighteen articles and eighteen apps were included in this study. Although many MS-related apps exist, only ten apps were identified in the systematic review to have published literature surrounding their use. Twelve articles looked at convergence between app-based testing and neurological exam. Overall, correlation between app v...
2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017
The vigor of a tree defines its ability to grow and is associated with its productivity ([1]; [2]... more The vigor of a tree defines its ability to grow and is associated with its productivity ([1]; [2]). The measurement of a tree's vigor makes it possible to evaluate its general state of health and its development over time. This task is traditionally performed by identifying defects affecting the vigor such as visible signs and symptoms on the tree, like fungi on the main stem, significant forks, cracks, wounds, competition and the percentage of living crown ([2]). The vigor and quality of trees are important factors used by forest managers in the application of silvicultural prescriptions at stand and landscape levels. The estimation of the level of risk of vigor loss in forest stands thus becomes a major and essential issue for the management and the improvement of the productivity of forests in a context of sustainable development.
Methods in Ecology and Evolution This article has been accepted for publication and undergone ful... more Methods in Ecology and Evolution This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as
This paper presents a new steganalysis method that uses a deep belief network (DBN) as a classifi... more This paper presents a new steganalysis method that uses a deep belief network (DBN) as a classifier for audio files. It has been tested on three steganographic techniques: StegHide, Hide4PGP and FreqSteg. The results were compared to two other existing robust steganalysis methods based on support vector machines (SVMs) and Gaussian mixture models (GMMs). Afterwards, another classification task aiming at identifying the type of steganographic applied or not to the speech signal was carried out. The results of this four-way classification show that in most cases, the proposed DBN-based steganalysis method gives higher classification rates than the two other steganalysis methods based on SVMs and GMMs.
IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium, 2008
ABSTRACT The present study develops a method for aboveground forest biomass mapping from Ikonos i... more ABSTRACT The present study develops a method for aboveground forest biomass mapping from Ikonos imagery and geospatial data. Reference biomass values by group of species were estimated using Ker's equations and inventory data from permanent sample plots (PEP) of 400 m2. A supervised classification of the Ikonos image, based on the maximum likelihood method presenting the five species groups inventoried in the field study, was carried out. Thereafter, various vegetation indices and texture parameters were extracted from the Ikonos image. The extracted Ikonos data were then combined with geospatial data at the same 1 m spatial resolution. Inventory plots biomass values estimated by group of species were used for the neural networks model (Multi-layer Perceptron) training with the backpropagation algorithm. Thereafter, biomass values for sample pixels generated randomly by group of species were predicted with the Multi-layer Perceptron. Then, sample pixels biomass values of each group were used to derive biomass values of other pixels of the same species group by interpolation with the ordinary kriging method using five different variogram models. The Gaussian variogram model yielded the best biomass estimates by comparison with reference biomass values, with percentages of residual errors ranging between 2,6 and 9,8% (absolute value) and percentages of RMSE (root mean square error) ranging between 17.2 and 61.1%.
This paper describes a method to compute dense correspondences (disparity map) between two images... more This paper describes a method to compute dense correspondences (disparity map) between two images in a stereo pair in case of significant displacements. In order to deal with large amplitude disparity fields, a multi-resolution technique called block match ing have been used in the past(1,2). Although quite successful, this technique cann ot easily cope with disparity discontinuities as it does
2005 ICSC Congress on Computational Intelligence Methods and Applications
A template-based technique for automatic recognition of birdsong syllables is presented. This tec... more A template-based technique for automatic recognition of birdsong syllables is presented. This technique combines time delay neural networks (TDNNs) with an autoregressive (AR) version of the backpropagation algorithm in order to improve the accuracy of bird species identification. The proposed neural network structure (AR-TDNN) has the advantage of dealing with a pattern classification of syllable alphabet and also of capturing
2012 11th International Conference on Information Science, Signal Processing and their Applications (ISSPA), 2012
Content-based indexing of images consists in extracting visual information from digital images (s... more Content-based indexing of images consists in extracting visual information from digital images (such as pixels, colors, objects, shapes, etc.), and can be performed automatically and fast by computers. This work focuses on color indexing of images. A statistical algorithm called Mean-Shift is used to modelize the color distributions of images as two-dimensional Gaussian kernels. The experiments have been performed on the COIL1 database and the results show that the proposed method compares well to other content-based retrieval methods.
2011 Seventh International Conference on Signal Image Technology & Internet-Based Systems, 2011
The topic of this paper is Content-Based Image Retrieval (CBIR) based on colors as a content imag... more The topic of this paper is Content-Based Image Retrieval (CBIR) based on colors as a content image descriptor. The tool we developed to that purpose modelizes the colors of an image as a set of 2D Gaussian distributions based on weighted color histograms. Then, given a reference image proposed by a user, the system can automatically classify the image and
2012 IEEE International Geoscience and Remote Sensing Symposium, 2012
This study proposes an approach which simultaneously uses spatial information and polarimetric da... more This study proposes an approach which simultaneously uses spatial information and polarimetric data from a RADARSAT-2 quad-polarization satellite image for forest tree species classification. The study area is near the Gounamitz River located in northwestern New Brunswick (Canada). After geometric correction of the image, two statistical models were used for the classification: (1) a Markov random fields model based on an initial segmentation provided by the K-means algorithm to account for the spatial statistical dependencies between adjacent sites; and (2) a K-distribution model with, as parameters, the covariance matrix containing all of the polarimetric information. The classification was optimized using the stochastic simulated annealing algorithm. Validation of the results was performed by comparison with field inventory measurements. Variation of the backscattering coefficient c° obtained for the RADARSAT-2 quad-polarization SAR image with incidence angles of 26 0 and 45 ° ranged from 1 and 3 dB for the different tree species. The results of average and overall accuracies of the classification were respectively 77.13% and 72.35% for the 26° incidence angle image compared to 81.47% and 79.12% for the 45°incidence angle.
2007 IEEE International Geoscience and Remote Sensing Symposium, 2007
ABSTRACT The present study develops an approach for the estimation of aboveground forest biomass ... more ABSTRACT The present study develops an approach for the estimation of aboveground forest biomass based on neural networks, using Ikonos satellite image data and multi-source geo-scientific data. Two methods of aboveground forest biomass estimation were compared: multiple regressions and the neural networks. Percentages of residual errors of the neural networks biomass estimates were lower than 1% for all groups of species, except for "intolerant hardwood" which had a percentage of 3.2% for the 9-18 cm DBH class. Percentages of residual errors of biomass estimates were higher with the quadratic multiple regression approach than with the neural networks, particularly for "intolerant hardwood" where a value of 51.41% was observed for the 19-40+ cm DBH class. Root mean square error values (RMSE) calculated from biomass estimates resulting from the neural networks approach were lower than those computed with estimates of the quadratic multiple regressions model, for all groups of species.
2016 IEEE Congress on Evolutionary Computation (CEC), 2016
This paper presents a new method to train Restricted Boltzmann Machines (RBMs) using Evolutionary... more This paper presents a new method to train Restricted Boltzmann Machines (RBMs) using Evolutionary Algorithms (EAs), where RBMs are used in the first step of a steganalysis tool for speech/audio files. The following EAs have been tested: Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Artificial Bees Colony (ABC) and Cat Swarm Optimization (CSO). Our method has been tested with three steganographic techniques: StegHide, Hide4PGP, and FreqSteg. A fourth technique combining the three steganographic methods has also been tested. The results are compared to the conventional contrastive divergence learning algorithm. All EAs outperform the contrastive divergence algorithm.
A lot of hype has recently been generated around deep learning, a group of artificial intelligenc... more A lot of hype has recently been generated around deep learning, a group of artificial intelligence approaches able to break accuracy records in pattern recognition. Over the course of just a few years, deep learning revolutionized several research fields such as bioinformatics or medicine. Yet such a surge of tools and knowledge is still in its infancy in ecology despite the ever-growing size and the complexity of ecological datasets. Here we performed a literature review of deep learning implementations in ecology to identify its benefits in most ecological disciplines, even in applied ecology, up to decision makers and conservationists alike. We also provide guidelines on useful resources and recommendations for ecologists to start adding deep learning to their toolkit. At a time when automatic monitoring of populations and ecosystems generates a vast amount of data that cannot be processed by humans anymore, deep learning could become a necessity in ecology.
Le radar a synthese d'ouverture (RSO) est un puissant outil de teledetection capable de fourn... more Le radar a synthese d'ouverture (RSO) est un puissant outil de teledetection capable de fournir des images a haute resolution de la surface terrestre par tout temps, de jour comme de nuit. Les images radar sont caracterisees par une forte granulation appelee speckle, inherente a la formation meme des donnees radar et qui rend extremement difficile leur interpretation automatique. L'objectif de cette these concerne la radargrammetrie ou stereovision radar, qui consiste a produire des modeles d'elevation de terrain a partir de paires d'images RSO. Pour cela il faut passer notamment par une phase-cle appelee appariement qui peut se faire a court terme en deux etapes : on effectue d'abord un filtrage du speckle, c'est-a-dire une estimation des reflectivites radar sous-jacentes de la surface imagee. Ceci permet alors d'appliquer aux images filtrees des methodes de traitement d'images eprouvees en imagerie optique. Cette approche en deux temps est cependant...
Uploads
Papers by Eric Hervet