Centre de documentation
scientifique

Ensemble methods for pedestrian detection in dense crowds / Jennifer Vandoni (2019)  

Public Titre : Ensemble methods for pedestrian detection in dense crowds Type de document : Thèse/HDR Auteurs : Jennifer Vandoni, Auteur ; Sylvie Le Hégarat-Mascle, Directeur de thèse Editeur : Paris-Orsay : Université de Paris 11 Paris-Sud Centre d'Orsay Année de publication : 2019 Importance : 182 p. Format : 21 x 30 cm Note générale : bibliographie Thèse de Doctorat de l'Université Paris-Saclay, Sciences et technologies de l’information et de la communication (STIC), Spécialité : Traitement du Signal et des Images Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] algorithme d'apprentissage [Termes IGN] apprentissage dirigé [Termes IGN] apprentissage profond [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] classification par séparateurs à vaste marge [Termes IGN] comportement [Termes IGN] densité de population [Termes IGN] détection de piéton [Termes IGN] données multicapteurs [Termes IGN] étalonnage [Termes IGN] fusion de données [Termes IGN] taxinomie [Termes IGN] théorie de Dempster-Shafer Index. décimale : THESE Thèses et HDR Résumé : (auteur) The interest surrounding the study of crowd phenomena spanned during the last decade across multiple fields, including computer vision, physics, sociology, simulation and visualization. There are different levels of granularity at which crowd studies can be performed, namely a finer microanalysis, aimed to detect and then track each pedestrian individually; and a coarser macro-analysis, aimed to model the crowd as a whole. One of the most difficult challenges when working with human crowds is that usual pedestrian detection methodologies do not scale well to the case where only heads are visible, for a number of reasons such as absence of background, high visual homogeneity, small size of the objects, and heavy occlusions. For this reason, most micro-analysis studies by means of pedestrian detection and tracking methodologies are performed in low to medium-density crowds, whereas macro-analysis through density estimation and people counting is more suited in presence of high-density crowds, where the exact position of each individual is not necessary. Nevertheless, in order to analyze specific events involving high-density crowds for monitoring the flow and preventing disasters such as stampedes, a complete understanding of the scene must be reached. This study deals with pedestrian detection in high-density crowds from a monocamera system, striving to obtain localized detections of all the individuals which are part of an extremely dense crowd. The detections can be then used both to obtain robust density estimation, and to initialize a tracking algorithm. In presence of difficult problems such as our application, supervised learning techniques are well suited. However, two different questions arise, namely which classifier is the most adapted for the considered environment, and which data to use to learn from. We cast the detection problem as a Multiple Classifier System (MCS), composed by two different ensembles of classifiers, the first one based on SVM (SVM-ensemble) and the second one based on CNN (CNN-ensemble), combined relying on the Belief Function Theory (BFT) designing a fusion method which is able to exploit their strengths for pixel-wise classification. SVM-ensemble is composed by several SVM detectors based on different gradient, texture and orientation descriptors, able to tackle the problem from different perspectives. BFT allows us to take into account the imprecision in addition to the uncertainty value provided by each classifier, which we consider coming from possible errors in the calibration procedure and from pixel neighbor’s heterogeneity in the image space due to the close resolution of the target (head) and descriptor respectively. However, scarcity of labeled data for specific dense crowd contexts reflects in the impossibility to easily obtain robust training and validation sets. By exploiting belief functions directly derived from the classifiers’ combination, we therefore propose an evidential Query-by-Committee (QBC) active learning algorithm to automatically select the most informative training samples. On the other side, we explore deep learning techniques by casting the problem as a segmentation task in presence of soft labels, with a fully convolutional network architecture designed to recover small objects (heads) thanks to a tailored use of dilated convolutions. In order to obtain a pixel-wise measure of reliability about the network’s predictions, we create a CNN-ensemble by means of dropout at inference time, and we combine the different obtained realizations in the context of BFT. To conclude, we show that the dense output map given by the MCS can be employed not only for pedestrian detection at microscopic level, but also to perform macroscopic analysis, bridging the gap between the two levels of granularity. We therefore finally focus our attention to people counting, proposing an evaluation method that can be applied at every scale, resulting to be more precise in the error and uncertainty evaluation (disregarding possible compensations) as well as more useful for the modeling community that could use it to improve and validate local density estimation. Note de contenu : 1- Crowd understanding 2- Supervised learning and classifier combination 3- SVM descriptors for pedestrian detection in high-density crowds 4- Taking into account imprecision with Belief Function Framework 5- Evidential QBC Active Learning 6- CNNs for pedestrian detection in high-density crowds 7- CNN-ensemble and evidential Multiple Classifier System 8- Density Estimation Conclusion Numéro de notice : 25704 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse française Note de thèse : Thèse de Doctorat : Traitement du Signal et des Images : Paris 11 : 2019 Organisme de stage : Systèmes et applications des technologies de l'information et de l'énergie (Paris) nature-HAL : Thèse DOI : sans En ligne : https://theses.hal.science/tel-02318892/document Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=94838

Use Markov random fields for automatic cloud-shadow detection on high resolution / Sylvie Le Hégarat-Mascle in ISPRS Journal of photogrammetry and remote sensing, vol 64 n° 4 (July - August 2009)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 64 n° 4 (July - August 2009) . - pp 351 - 366 Titre : Use Markov random fields for automatic cloud-shadow detection on high resolution Type de document : Article/Communication Auteurs : Sylvie Le Hégarat-Mascle, Auteur ; Cyrille André, Auteur Année de publication : 2009 Article en page(s) : pp 351 - 366 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] champ aléatoire de Markov [Termes IGN] connexité (graphes) [Termes IGN] détection automatique [Termes IGN] graphe [Termes IGN] image optique [Termes IGN] nuage [Termes IGN] ombre [Termes IGN] pixel [Termes IGN] processus ponctuel marqué Résumé : (Auteur) In this study, we propose an automatic detection algorithm for cloud/shadow on remote sensing optical images. It is based on physical properties of clouds and shadows, namely for a cloud and its associated shadow: both are connex objects of similar shape and area, and they are related by their relative locations. We show that these properties can be formalized using Markov Random Field (MRF) framework at two levels: one MRF over the pixel graph for connexity modelling, and one MRF over the graph of objects (clouds and shadows) for their relationship modelling. Then, we show that, practically, having performed an image pre-processing step (channel inter-calibration) specific to cloud detection, the local optimization of the proposed MRF models leads to a rather simple image processing algorithm involving only six parameters. Using a 39 image database, performance is shown and discussed, in particular in comparison with the Marked Point Process approach. Numéro de notice : A2009-294 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2008.12.007 En ligne : https://doi.org/10.1016/j.isprsjprs.2008.12.007 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=29924 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
081-09041	SL	Revue	Centre de documentation	Revues en salle	Disponible

Performance of change detection using remotely sensed data and evidential fusion: comparison of three cases of application / Sylvie Le Hégarat-Mascle in International Journal of Remote Sensing IJRS, vol 27 n°15-16 (August 2006)  

Public [article]  inInternational Journal of Remote Sensing IJRS > vol 27 n°15-16 (August 2006) . - pp 3515 - 3532 Titre : Performance of change detection using remotely sensed data and evidential fusion: comparison of three cases of application Type de document : Article/Communication Auteurs : Sylvie Le Hégarat-Mascle, Auteur ; R. Seltz, Auteur ; Laurence Hubert-Moy, Auteur ; Samuel Corgne, Auteur ; Nicolas Stach , Auteur Année de publication : 2006 Article en page(s) : pp 3515 - 3532 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] analyse comparative [Termes IGN] analyse texturale [Termes IGN] changement climatique [Termes IGN] classification de Dempster-Shafer [Termes IGN] cultures [Termes IGN] détection de changement [Termes IGN] forêt [Termes IGN] Pinus (genre) [Termes IGN] théorie de l'information Résumé : (Auteur) The detection of changes affecting continental surfaces has important applications in hydrological, meteorological and climatic modelling. Using remote sensing data, numerous change indices have already been proposed. Previous work showed the interest of combining several of these to improve change detection performance, using the Dempster–Shafer evidence theory framework. This study analyses the performance of different change indices and their combination in different cases of application: forest logging either in pine forest or in mixed forest, and winter vegetation cover of fields in intensive farming areas, in comparison to the forest fire case presented in previous work. The interest of indices derived from Information Theory, some of which are original, is shown. Numéro de notice : A2006-338 Affiliation des auteurs : IFN+Ext (1958-2011) Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1080/01431160500300255 Date de publication en ligne : 22/02/2007 En ligne : https://doi.org/10.1080/01431160500300255 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=28062 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
080-06081	RAB	Revue	Centre de documentation	En réserve L003	Disponible

Land covers change detection at coarse spatial scales based on iterative estimation and previous state information / Sylvie Le Hégarat-Mascle in Remote sensing of environment, vol 95 n° 4 (30/04/2005)  

Public [article]  inRemote sensing of environment > vol 95 n° 4 (30/04/2005) . - pp 464 - 479 Titre : Land covers change detection at coarse spatial scales based on iterative estimation and previous state information Type de document : Article/Communication Auteurs : Sylvie Le Hégarat-Mascle, Auteur ; Catherine Ottle, Auteur ; Christiane Guérin, Auteur Année de publication : 2005 Article en page(s) : pp 464 - 479 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] analyse diachronique [Termes IGN] bassin hydrographique [Termes IGN] chaîne de Markov [Termes IGN] détection de changement [Termes IGN] estimation statistique [Termes IGN] image à basse résolution [Termes IGN] image NOAA-AVHRR [Termes IGN] image SPOT-Végétation [Termes IGN] itération [Termes IGN] occupation du sol [Termes IGN] pixel [Termes IGN] Saône (rivière) [Termes IGN] sylviculture Résumé : (Auteur) This study focuses on the use of coarse spatial resolution (CR, pixel size about 1kM2) remote sensing data for land cover change detection and qualification. Assuming the linear mixing model for CR pixels, the problem is that both the multitemporal class feature and the pixel composition in terms of classes are unknown. The proposed algorithm is then based on the iterative alternate estimation of each unknown variable. At each iteration, the class features are estimated, thanks to the knowledge of the composition of so pixels, and then the pixel composition is re-estimated knowing the class features. The subset of known composition pixels is the sub of pixels where no change has occurred, i.e. the previous land cover map is still valid. It is derived automatically by removing at each iteration the pixels where the new composition estimation disagrees with the former one. Finally, for the final estimation of the pixel composition, a Markovian chain model is used to guide the solution, i.e. the previous land cover map is used as a 'reminder' 'memory' term. This approach has been first validated using simulated data with different spatial resolution ratios. Then, the detection of forest change with SPOT-VGT-S 10 has been considered as an actual application case. Finally, the method has been applied to change detection on the Val de Saône watershed between the 1980s and 2000. The results obtained from three coarse resolution series, NOAA/AVHRR, SPOT/VGT-S 10 and SPOT/VGT-P, have been compared. Numéro de notice : A2005-187 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.rse.2005.01.011 En ligne : https://doi.org/10.1016/j.rse.2005.01.011 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=27324 [article]

Automatic change detection by evidential fusion of change indices / Sylvie Le Hégarat-Mascle in Remote sensing of environment, vol 91 n° 3 (30/06/2004)  

Public [article]  inRemote sensing of environment > vol 91 n° 3 (30/06/2004) . - pp 390 - 404 Titre : Automatic change detection by evidential fusion of change indices Type de document : Article/Communication Auteurs : Sylvie Le Hégarat-Mascle, Auteur ; R. Seltzer, Auteur Année de publication : 2004 Article en page(s) : pp 390 - 404 Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] analyse comparative [Termes IGN] analyse texturale [Termes IGN] détection de changement [Termes IGN] dommage matériel [Termes IGN] fusion de données [Termes IGN] image optique [Termes IGN] incendie de forêt [Termes IGN] théorie de Dempster-Shafer [Termes IGN] valeur radiométrique Résumé : (Auteur) The detection of changes affecting continental surfaces has important applications in hydrological, meteorological, and climatic modeling. We propose a way to improve mono-index change detection by a fusion of multi-index change detection results. This fusion is performed in the framework of the Dempster-Shafer evidence theory, which is particularly suited to the representation of imprecision and ignorance at the "no change"/" change" class border. Depending on the change detection index considered, we also need to determine the class number and features. This is done using the a contrario theory approach rather than classical statistical tests. The proposed algorithm is applied to forest fire damage evaluation based on three popular change indices: normalized difference values, texture evolution, and mutual information (MI). We find that change index fusion is effective at reducing both false alarm and misdetection levels, due to the complementary nature of these indices. Numéro de notice : A2004-281 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.rse.2004.04.001 En ligne : https://doi.org/10.1016/j.rse.2004.04.001 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=26808 [article]

Soil moisture estimation from ERS-SAR data: toward an operational methodology / Sylvie Le Hégarat-Mascle in IEEE Transactions on geoscience and remote sensing, vol 40 n° 12 (December 2002)  

Permalink