Modélisation plausible d’objets 3D à partir de mesures multisensorielles L'objectif    à    long    terme    de    ce projet        vise        à        rapprocher davantage    les    environnements virtualisés     de     la     réalité     et     à mieux        les        adapter        aux applications    qui    requièrent    un haut     niveau     de     réalisme     en développant        des        modèles interactifs   3D   exacts   à   partir   de la      combinaison      de      mesures visuelles, tactiles et audio.  

Intelligence informatisée dans le

traitement de données complexes

Cette      direction      de      recherche      vise l’avancement             des             techniques intelligentes,   afin   d’assurer   un   traitement hautement      automatisé      et      l’utilisation complète,   efficace   et   en   temps   opportun des       données       dans       de       multiples applications.      Collaborations      avec      le Centre   de   recherche   IBM   sur   l’analytique et      la      performance      de      l’Université d’Ottawa      et      le      Service      de      Police d’Ottawa; Larus Technologies.

Saillance pour la modélisation et la

reconnaissance des formes

Ce   projet   porte   sur   le   développement   de méthodes          calculatoires          hybrides, combinant    les    méthodes    inspirées    de l'attention      visuelle      humaine      et      les algorithmes     pour     la     detection     de     la saillance    avec    les    méthodes    modernes de   traitement   d'images   et   des   données 3D.    
Quantification du mouvement pour l’interfaçage naturel homme-machine L’objectif   de   cette   direction   de   recherche   est   la quantification       et       la       reconnaissance       du mouvement     humain     en     capitalisant     sur     la capture   de   mouvements   sans   marqueurs   visant l’amélioration   de   la   performance   et   de   la   qualité des    interactions    homme-ordinateur    et    homme- machine. Modélisation et utilisation des environnements virtuels pour le traitement des phobies en psychologie Les    méthodes    développées    visent    l'ajustement approprié   du   niveau   de   détail   pour   la   modélisation des    objets    inclus    dans    des    applications    de    la réalité   virtuelle   pour   le   traitement   des   phobies   afin d'accroître      la      vitesse      d'affichage      dans      un environnement   virtuel   sans   pénaliser   le   réalisme.   Collaboration   avec   prof.   S.   Bouchard,   Laboratoire de Cyberpsychologie de l’UQO

Recherche

© Ana-Maria Cretu 2016

Ana-Maria Cretu

                                       ing. jr., Ph.D.
Les étudiant(e)s qui souhaitent poursuivre leurs études supérieures dans les domaines suivants: intelligence informatisée, analyse intelligente d’images, modélisation inspirée par la biologie, modélisation des objets rigides et déformables, réalité virtuelle, interfaces personne-machine, intelligence des systèmes, sont prié(e)s de contacter  Dr. Ana-Maria Cretu.

Informations

additionnelles

Programmes à l’UQO   Admission à l’UQO Support aux étudiant(e)s:      •  CRSNG      •  FRQNT        •  Bourses UQO   T.E.A. Oliveira (Doctorat) G. Rouhafzay (Doctorat) F. Audet (Doctorat) B. Tawbe (Doctorat) J.C. Davila Mesa* (Doctorat) T. Layadi (Doctorat) F. Hui* (Maîtrise) S. Filiatrault (Maîtrise) A. Mageau-Pétrin (Maîtrise) G. Plouffe* (Maîtrise) M. Chagnon-Forget (Maîtrise) N. Pedneault (Maîtrise) A. Quenneville (projet de synthése) T.G. Nkuzimana (stage) J.-M. Hébert (stage) Alumni: C. Viau* (Maîtrise) J. McCausland* (Maîtrise) M.I. Sina* (Maîtrise) Projets de synthèse: A. Quenneville F. Gorman N. Duchaine-Ritchot L. Couture-Niles A. Huot M. Chagnon-Forget J.-P. Gauthier L.-P. Fillion H. Monette-Thériault J. Tremblay-Gosselin Projets de fin d’études: P. Roy-Villeneuve V. Larivière P.-P. Chhay M. Cavrag V. D’Aoust M.-A. Charpentier S. Filiatrault S. Gauthier Stage de recherche: C. C. Diaw N. Bélanger L.H. Franc* T.T. Firmino de Lima (stage Mitacs Globalink) V. Bairaboina (stage Mitacs Globalink) P. Gauthier Q. Le Délas I. Filali* T. Level* Y.-F. Nassala* *co-supervision

Projets de recherche

This   project   explores   sensing   technologies   as   well   as   data   integration techniques   for   3D   deformable   object   modeling.   In   this   context,   a   data- driven     neural-network-based     model     for     capturing     implicitly     an d predicting    3D    deformations    of    a    soft    object’s    surface    subject    to external   forces   is   proposed.   Visual   data,   in   form   of   3D   point   clouds gathered   with   a   Kinect   sensor,   is   collected   over   an   object   while   forces are   exerted   by   means   of   the   probing   tip   of   a   force-torque   sensor.   A novel    approach    based    on    neural    gas    fitting    is    then    proposed    to describe    the    particularities    of    a    deformation    over    the    selectively simplified   3D   surface   of   the   object   without   requiring   knowledge   on   the object   material.   An   Elman   neural   network   is   finally   trained   to   predict the   mapping   between   the   measured   parameters   characterizing   the interaction with the obj ect and the change in the object shape due to the interaction. Support: Related publications: B.   Tawbe    and   A.-M.   Cretu,   “Acquisition   and   Prediction   of   Soft   Object   Shape   Using   a   Kinect   and   a   Force-Torque Sensor”, submitted to IEEE Trans. Instrumentation and Measurement, June 2016, NEW . T he   ill-defined   nature   of   tactile   information   turns   tactile sensors   into   th e   last   frontier   to   robots   that   can   handle everyday    objects    and    interact    with    humans    through contact.   To   overcome   this   frontier   many   design   aspects have       to       be       considered:       sensors       placement, electronic/mechanical   hardware,   methods   to   access   and acquire   signals,   calibration,   and   algorithms   to   process   and interpret   sensing   data   in   real   time. This   project   aims   at   the design    and    hardware    implementation    of    a    bio-inspired tactile   module   that   comprises   a   32   taxels   array   and   a   9   DOF   MARG   (Magnetic, Angular    Rate,    and    Gravity)    sensor,    a    flexible    compliant    structure    and    a    deep    pressure    sensor.    Making    use    of    a complementary   filter   the   orientation   of   the   shallow   sensors   can   be   estimated   and   the   force   applied   on   such   sensors   is conducted through the compliant structure to the deep pressure sensor. Related publications: T.E.A.   Oliveira , A.-M.   Cretu   and   E.M.   Petriu,   “Design   of   a   Multi-Modal   Bio-Inspired   Tactile   Module”,   submitted   to IEEE Trans. Instrumentation and Measurement,  June 2016, NEW . T.E.A.   Oliveira ,   V.P.   Fonseca,   A.-M.   Cretu,   E.M.   Petriu,   “Multi-Modal   Bio-Inspired   Tactile   Module”,   UOttawa Graduate   and   Research   Day,   University   of   Ottawa,   Mar.   2016   ( Research   Poster   Prize   in   Electrical   Engineering, first place; IEEE Research Poster Prize, first place ), NEW . Ce    projet    vise    le    développement    d’un    système    de    vision    périphérique    qui   permettra    de    surveiller    la    posture    du    corps    humain    et    de    quantifier    le mouvement   au   niveau   des   articulations.   Plus   précisément,   le   travail   consiste   à développer   des   solutions   logicielles   basées   sur   la   vision   pour   l'acquisition   et   la quantification   de   données   concernant   la   posture   humaine.   Il   s'agit   de   mesurer en   temps   réel   la   position   et   l'orientation   dans   l'espace   3D   des   articulations humaines    et    d'afficher        les    informations    concernant    la    quantification    des mouvements   exécutés   au   niveau   des   articulations.   Une   adaptation   est   proposée pour le contexte de la pédagogie d u pia no. Collaborateurs/ Financement: Publications: P.   Payeur,   J.   Beacon,   A.-M.   Cretu,   G.   M.   Nascimento,   G.   Comeau,   V. D'Aoust,   and   M.-A.   Charpentier ,   "Human   Gesture   Quantification:   an   Evaluation   Tool   for   Somatic   Training   and Piano   Peformance",   IEEE   Symp.   Haptic Audio-Visual   Environments   and   Games ,   pp.   100-105,   Dallas,   Texas,   US, Oct. 2014. S.    Gauthier ,    A.-M.    Cretu,    "Human    Movement    Quantification    using    Kinect    for    In-Home    Physical    Exercise Monitoring”,   IEEE   Int.   Conf.   Computational   Intelligence   and   Virtual   Environments   for   Measurement   Systems   and Applications , pp. 6-11, Ottawa, May 2014. One   of   the   tracks   of   the   project   aims   at   the   development   a natural    gesture     interface    that    tracks    and    recognizes    in real-time   static   and   dynamic   hand   gestures   of   a   user   based on   depth   data   collected   by   a   Kinect   sensor. A   novel   algorithm   is proposed   to   improve   the   scanning   time   in   order   to   identify   the first   pixel   on   the   hand   contour   within   this   space.   Starting   from this    pixel,    a    directional    search    algorithm    allows    for    the identification    of    the    entire    hand    contour.    The    K-curvature algorithm   is   then   employed   to   locate   the   fingertips   over   the contour,   and   dynamic   time   warping   s   used   to   select   gesture candidates   and   also   to   recognize   gestures   by   comparing   an observed   gesture   with   a   series   of   pre-recorded   reference   gestures.   Two   possible   applications   of   this   work   are   discussed and   evaluated:   one   for   interpretation   of   sign   digits   and   popular   gestures   for   a   friendlier   human-machine   interaction,   the other one for the natural control of a software interface. Another   track   of   the   project   aims   at   the   development   of   a   system   capable   to control    the   arm   movement   of   a   robot   by   mimicking   the   gestures   of   an actor   captured   by   a   markerless   vision   sensor.   The   Kinect   for   Xbox   is   used   to recuperate   angle   information   at   the   level   of   the   actor's   arm   and   an   interaction module   transforms   it   into   a   usable   format   for   real-time   robot   arm   control.   To avoid   self-collisions,   the   distance   between   the   two   arms   is   computed   in   real- time   and   the   motion   is   not   executed   if   this   distance   becomes   smaller   the twice   the   diameter   of   the   member.   A   software   architecture   is   proposed   and implemented   for   this   purpose. The   feasibility   of   our   approach   is   demonstrated on a NAO robot. Related publications:   G.   Plouffe ,   and A.-M.Cretu,   “Static   and   Dynamic   Hand   Gesture   Recognition   in   Depth   Data   Using   Dynamic   Time Warping“, IEEE Trans. Instrumentation and Measurement , vol. 65, no.2, pp. 305-316, Feb. 2016, NEW . G.   Plouffe ,   A.-M.   Cretu,   and   P.   Payeur,   “Natural   Human-Computer   Interaction   Using   Hand   Gestures”,   IEEE Symp. Haptic Audio-Visual Environments and Games , pp. 57-62, Ottawa, ON, Oct. 2015. S.   Filiatrault ,   and A.-M.   Cretu,   "Human Arm   Motion   Imitation   Using   a   Humanoid   Robot",   IEEE   Int.   Symp.   Robotic and Sensors Environments , pp. 31-36, Timisoara, Romania, 2014 . This    project    aims    at    proposing    automated    solutions    for    3D    object    modeling    at multiple resolutions in the context of virtual reality. An   original   solution,   based   on   an   unsupervised   neural   network,    is   proposed   to guide   the   creation   of   selectively   densified   meshes. A   neural   gas   network,   adapts   its nodes   during   training   to   capture   the   embedded   shape   of   the   object.   Regions   of interest   are   then   identified   as   areas   with higher   density   of   nodes   in   the   adapted neural   gas   map.   Meshes   at   different   level of    detail    for    an    object,    which    preserve these   regions   of   interest,   are   constructed by     adapting     a     classical     simplification algorithm.   The   simplification   process   will therefore   only   affect   the   regions   of   lower interest,   ensuring   that   the   characteristics of   an   object   are   preserved   even   at   lower resolutions.   A   novel   solution   based   on   learning   is   proposed   to   select   the number   of   faces   for   the   discrete   models   of   an   object   at   different   resolutions. Finally,   selectively   densified   object   meshes   are   incorporated   in   a   discrete level-of-detail method for presentation in virtual reality applications. Another   track   of   the   project   aims   at   developing   an   original   application   of   biologically-inspired     visual     attention     for      improved     perception-based modeling   of   3D   objects.   In   an   initial   step,   an   adapted   computational   model   of visual   attention   is   used   to   identify   areas   of   interest   over   the   3D   shape   of   an   object.   Points   of   interest   are   then   identified as    the    centroids    of    these    salient    regions    and    integrated,    along    with    their    immediate    n-neighbors,    and    using    a simplification algorithm, i nto a continuous distance-dependent level-of-details method. Support:   Publications:   M.    Chagnon-Forget ,    A.-M.    Cretu,    and    S.    Bouchard,    "Visual-Attention    Based    Interest    Point    Detection    for Perceptually-Improved 3D Object Modeling", submitted to ACM Trans. Applied Perception , NEW . A.-M.   Cretu,   M.   Chagnon-Forget    and   P.   Payeur,   “Selectively   Densified   3D   Object   Modeling   Based   on   Regions   of Interest   Using   Neural   Gas   Networks”,   accepted   for   publication,   Soft   Computing,   2016,   doi:   10.1007/s00500-016- 2132-z,   NEW . H.   Monette-Thériault,   A.-M.   Cretu,   and   P.   Payeur,   "3D   Object   Modeling   with   Neural   Gas   Based   Selective Densification   of   Surface   Meshes”,   IEEE   Int.   Conf.   Systems,   Man,   and   Cybernetics,    pp.   1373-1378,   San   Diego, US, 2014. Virtual   reality   has   already   been   successfully   used   as   a   therapeutic   tool   for   the    treatment   of various   phobias.   Due   to   advances   in   the   3D   graphics   and   in   the   computing   power,   the   real- time   visual   rendering   of   a   virtual   world   poses   no   significant   problems   nowadays.   However, the   haptic   interaction   with   such   environments   remains   a   challenge.   This   paper   explores   the haptic   interaction   with   a   dedicated   virtual   environment   in   spider   phobia   treatment   to   elicit disgust,   as   changes   in   fear   and   in   disgust   were   shown   to   be   highly   associated   with   the observed    decline    in    arachnophobic    symptoms.    A    dedicated    virtual    environment    is programmed   within   which   a   Novint   Falcon   haptic   device   is   used   for   the   interaction   with   a virtual spider.   Collaborators: Related publications:   M.   Laforest,   S.   Bouchard,   A.-M.   Cretu,   and   O.   Mesly,   “Inducing   an   Anxiety   Response   Using   a   Contaminated Virtual   Environment:   Validation   of   a   Therapeutic   Tool   for   Obsessive-Compulsive   Disorder”,   submitted   to   Frontiers in ICT (Information and Communication Technologies): Virtual Environments, Mar. 2016, NEW. M.   Cavrag,    G.   Larivière, A.-M.   Cretu,   and   S.   Bouchard,   "Interaction   with   Virtual   Spiders   for   Eliciting   Disgust   in   the Treatment   of   Phobias",   IEEE   Symp.   Haptic Audio-Visual   Environments   and   Games ,   pp.   29-34,   Dallas, Texas,   US, Oct. 2014. The long term objective of this research project is to attempt to establish reliable associations between traffic data and accident occurrence. These associations should help a proactive approach for accident management by an automatic detection of prone locations and conditions for accidents. Collaborators/ Support: Related publications:   A.-M.   Cretu,   A.   Mageau-Pétrin   and   C.   Hopgood,   “Traffic   Accident   Modeling   and   Prediction”,   Ottawa   Police Services Research Project, Technical report, May. 2016, 33 pages, NEW. A.-M.   Cretu,   and   A.   Mageau-Pétrin,   “Intelligence-led   traffic   enforcement:   Data   mining   techniques   for   accident prediction”, Ottawa Police Services Research Project, Technical report, Jan. 2016, 159 slides, NEW. The   objective   of   this   project   is   to   study   the   impact   of   the   use   of    the images/avatars for the detection of communities in social networks. Support: Recent    years    have    witnessed    an    increasing    interest    into    intelligent   autonomous   robots   able   to   execute   tasks   in   unknown   or   only   partially known   environments.   One   important   problem   to   deal   with   in   this   context is    the    capability    of    the    robot    to    autonomously    sense    its    immediate environment   and   understand   its   contents   in   order   to   be   able   to   navigate safely   and   perform   its   tasks   without   human   intervention.   The   objective   of this    project    is    to    extend    these    capabilities    by    a    fast    and    automated generation   of   an   environment   map,   associated   with   the   recognition   of   the objects   contained   in   it.   These   will   serve   later   for   robot   path   planning   and the execution of predetermined actions on the recognized objects. As     recent     research     has     demonstrated     that     depth     cues     have     a considerable   impact   on   human   fixations,   and   that   salient   maps   can   be improved   by   incorp orating   depth   information,   a   part   of   this   project   aims at    the    exploitation    of    depth    information    to    improve    segmentation    for salient object detection..  Collaborators/ Support: Related publications:   F.   Audet,    A.-M.   Cretu,   and   M.S.   Allili,   “Salient   Object   Segmentation   in   RGB-D   Data”,   submitted   to   Int.   Symp. Visual Computing , Las Vegas, Dec. 2016,  NEW. L'objectif   principal   de   ce   projet   est   le   développement   d'une méthode    efficace    et    automatisée    pour    l'identification    des édifices    dans    des    images    aériennes,    en    utilisant    des approches   modernes   pour   la   segmentation   d'images,   des méthodes   inspirées   de   l’attention   visuelle   et   des   techniques d'apprentissage automatisé.

Financement:

Publications: J.      Tremblay-Gosselin       and      A.-M.      Cretu,      "A Supervised Training   and   Learning   Method   for   Building Identification   in   Remotely   Sensed   Images",   IEEE   Int.   Symp.   Robotic   and   Sensors   Environments ,   Washington, US, 21-23 Oct. 2013, pp.73-78. A.-M.   Cretu,   and   P.   Payeur,   "Building   Detection   in   Aerial   Images   Based   on   Watershed   and   Visual   Attention Feature Descriptors", Canadian Conf.  Computer and Robot Vision , pp. 265-272, Regina, 2013. A.-M.   Cretu,   "Evolving   Sensor   System   Environments   with   Visual   Attention:   An   Experimental   Exploration",   IEEE Symp. Robotic and Sensors Environments , pp. 97-102, Magdeburg, Germany, 16-18 Nov. 2012 (invited talk) . This   project   explores   the   capabilities   of   computational   intelligence   for   fault   detection   in   sensor   networks.   Neural-network approaches   are   exploited   in   order   to   provide   a   general   solution   covering   typical   sensor   faults   and   to   replace   complex sets   of   individual   detection   methods.   For   this   purpose,   an   appropriate   set   of   fault   relevant   features   is   identified   in   a   first step.   A   generic   neural-network   structure   and   learning   strategy   is   then   chosen   and   adapted   for   detecting   multiple   fault types.   Afterwards   the   approach   is   applied   on   a   common   used   sensor   system   and   evaluated   with   deterministic    fault injections. Collaborators: Related publications : G.   Jager,   S.   Zug,   T.   Brade,   A.   Dietrich,   C.   Steup,   C.   Moewes,   A.-M.   Cretu,   “Assessing   Neural   Networks   for Sensor   Fault   Detection”,   IEEE   Int.   Conf.   Computational   Intelligence   and   Virtual   Environments   for   Measurement Systems and Applications , pp. 70-75,  Ottawa, May 2014 . Drawing     inspiration     from     the     significantly     superior performance   of   humans   to   extract   and   interpret   visual information,     the     proposed     research     transposes     the visual    attention    mechanisms    into    biologically-inspired computational   systems   to   develop   new   techniques   and computational    resources    capable    to    interpret    complex images     with     large     variations     in     their     content     and characteristics.     The     project     is     dedicated     to     the exploration,   implementation,   refinement   and   validation   of a   series   of   biologically-inspired   computational   models   for attention    and    object    recognition    dedicated    to    specific tasks,   such   as   building   detection   in   the   context   of   geo- imaging. Collaborators/Suppor t: Related publications: A.-M.   Cretu,   and   P.   Payeur,   "Visual   Attention   Model   with   Context   Learning   for   Building   Detection   in   Satellite Images", Int. Journal Smart Sensing and Intelligent Systems , vol. 5, no.4, pp. 742-766, Dec. 2012. M.    I.    Sina ,    P.    Payeur,    and   A.-M.    Cretu,    "Object    Recognition    on    Satellite    Images    with    Biologically-Inspired Computational   Approaches",   IEEE   Int.   Symp.   Applied   Computational   Intelligence   and   Informatics ,   pp.   81-86, Timisoara, Romania, May 2012. M.   I.   Sina,    A.-M.   Cretu,   and   P.   Payeur,   “Biological   Visual   Attention   Guided   Automatic   Image   Segmentation   with Application   in   Satellite   Imaging”,   Proc.   IS&T/SPIE   Electronic   Imaging   Conference,    Human   Vision   and   Electronic Imaging Track, vol. 8291, Burlingame, California, USA, Jan. 2012. The    automated    servicing    of vehicles     is     becoming     more and   more   a   reality   in   today’s world.           While           certain operations,      such      as      car washing,   require   only   a   rough model    of    the    surface    of    a vehicle,   other   operations,   such as    changing    of    a    wheel    or filling    the    gas    tank,    require correct   localization   of   the   different   parts   of   the   vehicle   on   which   operations   are   to   be   performed.   The   proposed   image- based   approach   to   roughly   localize   vehicle   parts   over   the   surface   of   a   vehicle   with   a   bounding   box   approach   is   based   on a   model   of   human   visual   attention.   The   proposed   method   is   automatically   adapted   for   different   views   of   a   vehicle   and obtains   average   localization   rates   for   different   vehicle   parts   of   over   95%   for   a   dataset   of   120   vehicles   belonging   to   three categories,   namely   sedan,   SUV   and   wagon   and   allows,   with   the   addition   of   the   active   contour   models,   for   a   more complete and accurate description of vehicle parts contours than other state-of-the-art solutions. Collaborators/Support: Related publications: D.   Nakhaeinia,   P.   Payeur,   A.   Chávez-Aragón,   A.-M.   Cretu,   R.   Laganière,   and   R.   Macknojia,   "Surface   Following with   an   RGB-D   Vision-Guided   Robotic   System   for   Vehicle   Security   Screening",   soumis   à   IEEE   Trans. Automation Science and Engineering , Aug. 2014 , NEW . A.-M.   Cretu,   P.   Payeur   and   R.   Laganière,   "An Application   of   a   Bio-Inspired   Visual Attention   Model   Guided   for   the Localization of Vehicle Parts", Applied Soft Computing , vol. 31. pp. 369-380, 2015, NEW . R.   Fareh,   P.   Payeur,   D.   Nakhaeinia,   R.   Macknojia, A.   Chávez-Aragón, A.-M.   Cretu,   P.   Laferrière,   R.   Laganière.   R. Toledo,    "An    Integrated    Vision-Guided    Robotic    System    for    Rapid    Vehicle    Inspection",    IEEE    Int.    Systems Conference , Ottawa, Mar. 2014 ( Best Paper Award ), pp. 446-451. A.-M.   Cretu   and   P.   Payeur,   "Image-Based   Localization   of   Vehicle   Parts   Guided   by   Visual   Attention",   Proc.   IEEE Int. Conf. Instrumentation and Measurement Technology,  pp. 533-538, Graz, Austria, May 2012. The    continuous    rise    in    the    amount    of    vehicles    in circulation       brings       an       increasing       need       for automatically     and     efficiently     recognizing     vehicle categories      for      multiple      applications      such      as optimizing   available   parking   spaces,   balancing   ferry space,   perceiving   highway   toll,   planning   infrastructure and    managing    traffic,    or    servicing    vehicles.    This project   describes   the   design   and   implementation   of   a vehicle   classification   system   using   a   set   of   images collected   from   6   views.   The   proposed   computational system        combines        human        visual        attention mechanisms   to   identify   a   set   of   salient   discriminative features    and    a    series    of    binary    support    vector machines to achieve fast automated classification. Collaborators/Support: Related publications: A.-M.    Cretu,    and    P.    Payeur,    "Biologically- Inspired     Visual     Attention     Features     for     a Vehicle   Classification Task",   Int.   Journal   Smart Sensing   and   Intelligent   Systems ,   vol.   4,   no.   3, pp. 402-423, Sep. 2011. A.-M.   Cretu,   P.   Payeur,   and   R.   Laganière,   "Salient   Features   Based   on   Visual   Attention   for   Multi-View   Vehicle Classification",   IEEE   Int.   Conf.   Computational   Intelligence   for   Measurement   Systems   and Applications ,   pp.   64-69, Ottawa, Canada, Sep. 2011. This    project    addresses    the    problem    of    node    selection    using    computational intelligence   techniques   for   a   risk-aware   robotic   sensor   networks   applied   to   critical infrastructure   protection. The   goal   is   to   maintain   a   secure   perimeter   around   a   critical infrastructure,   which   is   best   maintained   by   detecting   high-risk   network   events   and mitigate them through a response involving the most suitable robotic assets. Collaborators: Related publications: J.    McCausland,     R.   Abielmona,    R.    Falcon,   A.-M.    Cretu    and    E.M.    Petriu,    "On    the    Role    of    Multi-Objective Optimization   in   Risk   Mitigation   for   Critical   Infrastructures   with   Robotic   Sensor   Networks",   ACM   Conference Companion on Genetic and Evolutionary Computation , Vancouver, Canada, pp. 1269-1276 , 2014, NEW . J.   McCausland ,   R.   Abielmona,   A.   -M.   Cretu,   Rafael   Falcon,   and   E.M.   Petriu,   “A   Proactive   Risk-Aware   Robotic Sensor   Network   for   Critical   Infrastructure   Protection”,   poster   session,   2013   Annual   Research   Review   Healthcare Support through Information Technology Enhancements (hSITE) , Montreal, QC, 18 Nov. 2013,   J.   McCausland ,   R.   Abielmona,   R.   Falcon,   A.-M.   Cretu   and   E.M.   Petriu,   "Auction-Based   Node   Selection   of Optimal   and   Concurrent   Responses   for   a   Risk-Aware   Robotic   Sensor   Network",   IEEE   Int.   Symp.   Robotic   and Sensors Environments , Washington, US, 21-23 Oct. 2013, pp. 136-141 . The    project    discusses    the    design    and implementation        of        an        automated framework    that    provides    the    necessary information   to   the   controller   of   a   robotic hand    to    ensure    safe    model-based    3D deformable           object           manipulation. Measurements      corresponding      to      the interaction     force     at     the     level     of     the fingertips     and     to     the     position     of     the fingertips   of   a   three-fingered   robotic   hand are    associated    with    the    contours    of    a deformed    object    tracked    in    a    series    of images   using   neural-network   approaches.   The   resulting   model   not   only   captures   the   behavior   of   the   object   but   is   also able   to   predict   its   behavior   for   previously   unseen   interactions   without   any   assumption   on   the   object’s   material.   Such models   allow   the   controller   of   the   robotic   hand   to   achieve   better   controlled   grasp   and   more   elaborate   manipulation capabilities. Related publications: A.-M.   Cretu,   P.   Payeur,   and   E.M.   Petriu,   “Soft   Object   Deformation   Monitoring   and   Learning   for   Model-Based Robotic Hand Manipulation”, IEEE Trans. Systems, Man and Cybernetics - Part B , vol. 42, no. 3, Jun. 2012. A.-M.   Cretu,   P.   Payeur,   and   E.M.   Petriu,   “Learning   and   Prediction   of   Soft   Object   Deformation   using   Visual Analysis   of   Robot   Interaction”,      Proc.   Int.   Symp.   Visual   Computing , ISVC2010,   Las   Vegas,   Nevada,   USA,   G. Bebis et al. (Eds), LNCS 6454, pp. 232-241, Springer, 2010.     F.   Khalil,   P.   Payeur, and   A.-M.   Cretu,      “Integrated   Multisensory   Robotic   Hand   System   for   Deformable   Object Manipulation”, Proc. Int. Conf. Robotics and Applications , pp. 159-166, Cambridge, Massachusetts, US, 2010. A.-M.   Cretu,   P.   Payeur,   E.   M.   Petriu   and   F.   Khalil,   “Estimation   of   Deformable   Object   Properties   from   Visual   Data and   Robotic   Hand   Interaction   Measurements   for   Virtualized   Reality   Applications   ”,   Proc.   IEEE   Int.   Symp.   Haptic Audio Visual Environments and Their Applications , pp.168-173, Phoenix, AZ,  Oct. 2010.      A.-M.   Cretu,   P.   Payeur,   E.   M.   Petriu   and   F.   Khalil,   “Deformable   Object   Segmentation   and   Contour   Tracking   in Image   Sequences   Using   Unsupervised   Networks”,   Proc.   Canadian   Conf.   Computer   and   Robot   Vision ,   pp.   277- 284, Ottawa, Canada,  May 2010. This       project       explores       some aspects    of    intelligent    sensing    for advanced      robotic      applications, with      the      main      objective      of designing    innovative    approaches for   automatic   selection   of   regions of   observation   for   fixed   and   mobile sensors    to    collect    only    relevant measurements      without      human guidance.    The    proposed    neural gas      network      solution      selects regions     of     interest     for     further sampling   from   a   cloud   of   sparsely collected   3D   measurements.   The technique   automatically   determines bounded   areas   where   sensing   is   required   at   higher   resolution   to   accurately   map   3D   surfaces.   Therefore   it   provides significant    benefits    over    brute    force    strategies    as    scanning    time    is    reduced    and    the    size    of    the    dataset    is    kept manageable. Collaborators/Support: Related publications: P.   Payeur,   P.   Curtis,   A.-M.   Cretu,   "Computational   Methods   for   Selective   Acquisition   of   Depth   Measurements:   an Experimental   Evaluation",   Int.   Conf.      Advanced   Concepts   for   Intelligent   Vision   Systems ,   Poznan,   Poland,   J. Blanc-Talon et al. (Eds.) LNCS 8192, pp. 389-401, 2013. P.   Payeur,   P.   Curtis,   A.-M.   Cretu,   "Computational   Methods   for   Selective   Acquisition   of   Depth   Measurements   in Machine Perception", IEEE Int. Conf. Systems, Man, and Cybernetics,  Manchester, UK, 2013 pp.876-881. A.-M.   Cretu,   P.   Payeur,   and   E.M.   Petriu,   “Selective   Range   Data   Acquisition   Driven   by   Neural   Gas   Networks”, IEEE Trans. Instrumentation and Measurement , vol. 58, no. 6, pp. 2634-2642, 2009. A.-M.   Cretu,   P.   Payeur,   and   E.M.   Petriu,   “Neural   Gas   and   Growing   Neural   Gas   Networks   for   Selective   3D Sensing: A Comparative Study”, Sensors & Transducers , vol. 5, pp. 119-134, 2009. A.-M.   Cretu,   P.   Payeur,   and   E.M.   Petriu,   “Selective   Tactile   Data   Acquisition   on   3D   Deformable   Objects   for Virtualized   Reality   Applications”,   Proc.   IEEE   Int.   Workshop   Computational   Intelligence   in   Virtual   Environments , pp.14-19, Nashville, TN, USA, Apr. 2009. A.-M.   Cretu,   E.M.   Petriu,   and   P.   Payeur,   “Growing   Neural   Gas   Networks   for   Selective   3D   Scanning”,   Proc.   IEEE Int. Workshop on Robotic and Sensors Environments , pp. 108-113, Ottawa, Canada, Oct. 2008. A.-M.   Cretu,   P.   Payeur,   and   E.M.   Petriu,   “Selective   Vision   Sensing   Based   on   Neural   Gas   Network”,   Proc.   IEEE Int.   Conf.   Instrumentation   and   Measurement   Technology,    pp.   478-483,   Vancouver,   Canada,   May   2008   (Best Student Paper Award).   This    work    presents    a    general-purpose scheme   for   measuring,   constructing   and representing      geometric      and      elastic behavior   of   deformable   objects   without   a priori   knowledge   on   the   shape   and   the material   that   the   objects   under   study   are made     of.     The     proposed     solution     is       based   on   an   advantageous   combination of   neural   network   architectures   and   an original   force-deformation   measurement procedure.    An    innovative    non-uniform selective      data      acquisition      algorithm based       on       self-organizing       neural architectures    (namely    neural    gas    and growing    neural    gas)    is    developed    to selectively   and   iteratively   identify   regions of   interest   and   guide   the   acquisition   of   data   only   on   those   points   that   are   relevant   for   both   the   geometric   model   and   the mapping   of   the   elastic   behavior,   starting   from   a   sparse   point-cloud   of   an   object.   Multi-resolution   object   models   are obtained   using   the   initial   sparse   model   or   the   (growing   or)   neural   gas   map   if   a   more   compressed   model   is   desired,   and augmenting   it   with   the   higher   resolution   measurements   selectively   collected   over   the   regions   of   interest.   A   feedforward neural   network   is   then   employed   to   capture   the   complex   relationship   between   an   applied   force,   its   magnitude,   its   angle of   application   and   its   point   of   interaction, the    object    pose    and    the    deformation stage   of   the   object   on   one   side,   and   the object     surface     deformation     for     each region   with   similar   geometric   and   elastic behavior     on     the     other     side.     The proposed   framework   works   directly   from raw    range    data    and    obtains    compact point-based    models.    It    can    deal    with different          types          of          materials, distinguishes      between      the      different stages   of   deformation   of   an   object   and models       homogeneous       and       non- homogeneous objects as well. It also offers the desired degree of control to the user. Support: Related publication: A.M.   Cretu,   “Experimental   Data   Acquisition   and   Modeling   of   3D   Objects   using   Neural   Networks”,   Ph.D.   Thesis, University of Ottawa, 2009. This    project    presents    a    critical    comparison between    three    neural    architectures    for    3D object    representation    in    terms    of    purpose, computational   cost,   complexity,   conformance and   convenience,   ease   of   manipulation   and potential    uses    in    the    context    of    virtualized reality.     Starting     from     a     pointcloud     that embeds    the    shape    of    the    object    to    be modeled,     a     volumetric     representation     is obtained    using    a    multilayered    feedforward neural   network   or   a   surface   representation using   either   the   self-organizing   map   or   the neural     gas     network.     The     representation provided   by   the   neural   networks   is   simple, compact    and    accurate.   The    models    can    be easily    transformed    in    size,    position    (affine    transformations)    and    shape    (deformation).    Some    potential    uses    of    the presented   architectures   in   the   context   of   virtualized   reality   are   for   the   modeling   of   set   operations   and   object   morphing, for the detection of objects collision and for object recognition, object motion estimation and segmentation. Related publications: A.-M.   Cretu,   “Neural   Network   Modeling   of   3D   Objects   for   Virtualized   Reality   Applications”,   M.A.Sc.   Thesis, University of Ottawa, 2003. A.-M.   Cretu   and   E.M.   Petriu,   “Neural   Network-Based Adaptive   Sampling   of   3D   Object   Surface   Elastic   Properties”, IEEE Trans. Instrumentation and Measurement , vol. 55, no. 2, pp. 483-492, 2006. A.-M.   Cretu,   E.M.   Petriu,   and   G.G.   Patry,   “Neural-Network   Based   Models   of   3D   Objects   for   Virtualized   Reality: A Comparative Study”, IEEE Trans. Instrumentation and Measurement , vol. 55, no. 1, pp. 99-111, 2006. A.-M.   Cretu,   E.M.   Petriu,   and   G.G.   Patry,   “A   Comparison   of   Neural   Network   Architectures   for   the   Geometric Modeling   of   3D   Objects”,   Proc.   Conf.   Computational   Intelligence   for   Measurement   Systems   and Applications ,   pp. 155-160, Boston, USA, Jul. 2004. A.-M.   Cretu,   E.M.   Petriu,   and   G.G.Patry,   “Neural   Network Architecture   for   3D   Object   Representation”,   Proc.    IEEE Int.   Workshop   on   Haptic,   Audio   and   Visual   Environments   and   Their   Applications ,   pp.   31-36,   Ottawa,   Canada, Sep. 2003. Controlling    robotic    interventions    on    small    devices    creates important    challenges    on    the    sensing    stage    as    resolution limitations    of    non-contact    sensors    are    rapidly    reached.    The integration   of   haptic   sensors   to   refine   information   provided   by vision   sensors   appears   as   a   very   promising   approach   in   the development   of   autonomous   robotic   systems   as   it   reproduces the   multiplicity   of   sensing   sources   used   by   humans. This   project proposes   an   intelligent   multimodal   sensor   system   developed   to enhance   the   haptic-control   of   robotic   manipulations   of   small   3D objects. The   proposed   system   combines   a   16x16   array   of   Force Sensing     Resistor     (FSR)     elements     to     refine     3D     shape measurements   in   selected   areas   previously   monitored   with   a laser    range    finder.    Using    the    integrated    technologies,    the sensor    system    is    able    to    recognize    small-size    objects    that cannot   be   accurately   differentiated   through   range   measurements and   provides   an   estimate   of   the   objects   orientation.   Characteristics   of   the   system   are   demonstrated   in   the   context   of   a robotic intervention that requires fine objects to be localized and identified for their shape and orientation. Related publications: P.   Payeur,   C.   Pasca,   A.-M.   Cretu,   and   E.M.   Petriu,   “Intelligent   Haptic   Sensor   System   for   Robotic   Manipulation”, IEEE Trans. Instrumentation and Measurement , vol. 54, no. 4, pp. 1583-1592, 2005. C.   Pasca,   P.   Payeur,   E.M.   Petriu, A.-M.   Cretu,   “Intelligent   Haptic   Sensor   System   for   Robotic   Manipulation”,   Proc. IEEE Int. Conf. Instrumentation and Measurement Technology,  pp. 279-284, Italy, 2004. E.   M.   Petriu,   S.K.S.   Yeung,   S.R.   Das, A.-M.   Cretu,   and   H.J.W.   Spoelder,   “Robotic   Tactile   Recognition   of   Pseudo- Random Encoded Objects”, IEEE Trans. Instrumentation and Measurement,  vol. 53, no. 5, pp. 1425-1432, 2004. The    inherent    complex    nature of    the    wastewater    treatment process,    the    lack    of    proper knowledge   and   description   of the   biological   phenomena   and the   large   fluctuations   in   time   of the      numerous      parameters implied         (flow-rates         and nutrients   loadings)   makes   the automatic       control       of       a wastewater            plant            a complicated        and        difficult problem   to   tackle.   The   use   of online   sensors   for   continuous measurement    of    wastewater components      is      prone      to unpredictable         breakdowns. Low   nutrient   conditions   in   a   wastewater   treatment   plant   may   cause   the   failure   of   the   effective   nitrogen   and   phosphorus removal   for   a   considerable   period.   This   project   proposes   neural-network   modeling   approaches   for   the   model-based control of a wastewater treatment plant in terms of air flow-rate and municipal wastewater components. Related publications: K-Y.   Ko,   G.G.   Patry, A.-M.   Cretu,   and   E.M.   Petriu,   “Neural   Network   Model   of   Municipal   Wastewater   Components for    a    Wastewater    Treatment    Plant”,    Proc.    IEEE    Int.    Workshop    on    Advanced    Environmental    Sensing    and Monitoring Techniques , pp. 23-28, Como, Italy, Jul. 2003. K.-Y.   Ko,   G.G.   Patry,   A.-M.   Cretu,   E.M.   Petriu,   “Neural   Network   Model   for   Wastewater   Treatment   Plant   Control”, Proc.    IEEE    Int.    Workshop    on    Soft    Computing    Techniques    in    Instrumentation,    Measurement    and    Related Applications , pp. 38-43, May 2003.

Neuro-inspired computational intelligence for geo-imaging systems

Futur(e)s Étudiant(e)s
Étudiants

Localization of vehicle parts guided by visual attention

Salient features for image-based vehicle classification

Risk-aware wireless sensor network for critical infrastructure protection

Deformable object tracking and modeling for robotic hand manipulation 

Selective range data acquisition

Experimental data acquisition and modeling of 3D deformable objects using neural networks

Neural network modeling of 3D objects for virtualized reality applications

Pattern classification and recognition from tactile data

Neural network models for wastewater treatment plant control

Courtesy Larus Technologies

Quantification du mouvement humain

Computational intelligence in sensor environments

Perceptually-improved multi-resolution 3D object modeling

Apprentissage informatisé pour la reconnaissance des formes dans des images aériennes

Community detection in social networks using images

3D environment mapping and object recognition in RGBD data

Virtual environments for phobia treatment

Sensor systems for multisensory data acquisition

Multimodal biologically-inspired tactile sensing

Courtesy of  the Cyberpsychology lab UQO

Intelligence-led traffic enforcement

Motion tracking and imitation for natural human-machine interaction