<!DOCTYPE html>
<html>
  <head>
    <meta http-equiv="content-type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <div aria-live="assertive" id="magicdomid25" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">The
        FOX team from the CRIStAL laboratory (UMR CNRS), Lille France
        and the PR team from the MIS Laboratory, Amiens France are
        looking to recruit a joint PhD student starting in October 2024
        on the following subject: </span><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g b"><b>EventSpike
          - Async</b></span><span
        class="author-a-z71z3z80z3z81zz89zmz82zz71zfz74zz82zz73z7x3 b"><b>h</b></span><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g b"><b>ronous
          computer vision from event camera</b></span><span
        class="author-a-z71z3z80z3z81zz89zmz82zz71zfz74zz82zz73z7x3 b"><b>s</b></span></div>
    <div aria-live="assertive" id="magicdomid26" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> </span></div>
    <div aria-live="assertive" id="magicdomid27" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">Abstract:
        Video analysis is one of the fundamental tasks in computer
        vision. The dominant approach is based on deep neural networks
        applied to RGB images. These models have disadvantages such as:
        a) the need of large quantities of annotated data, which
        requires significant human work; b) the significant
        computational and therefore energy cost of these approaches; and
        c) redundancy in terms of visual information between two
        successive images. Spiking neural networks can offer a solution
        to these problems, through the use of unsupervised learning
        rules inspired by biological learning and the possibility of
        implementing them on ultra-low energy hardware components. .
        Event cameras that only communicate changes in light intensity
        are positioned as an alternative for capturing a scene when
        efficient processing on hardware with low computing capabilities
        is required. The objective of this thesis is to offer a joint
        response by proposing weakly supervised learning methods based
        on spiking learning mechanisms which will directly exploit the
        flow of impulses generated by an event camera. We are targeting
        applications around autonomous driving, such as the detection of
        moving vehicles or the recognition of information conveyed by
        large displays along the road infrastructure.</span></div>
    <div aria-live="assertive" id="magicdomid28" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> </span></div>
    <div aria-live="assertive" id="magicdomid29" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">The
        objective of this thesis is to develop new models of spiking
        neural networks (SNN) capable of directly processing visual
        information in the form of spike trains. The proposed models
        must be validated experimentally on dynamic vision databases,
        following standard protocols and best practices.</span></div>
    <div aria-live="assertive" id="magicdomid30" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> </span></div>
    <div aria-live="assertive" id="magicdomid31" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">Candidates
        must hold a Master degree (or an equivalent degree) in Computer
        Science, Statistics, Applied Mathematics or a related field.
        Experience in one or more of the following is a plus:</span></div>
    <div aria-live="assertive" id="magicdomid32" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> •
        image processing, computer vision;</span></div>
    <div aria-live="assertive" id="magicdomid33" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> •
        machine learning;</span></div>
    <div aria-live="assertive" id="magicdomid34" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> •
        bio-inspired computing;</span></div>
    <div aria-live="assertive" id="magicdomid35" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> •
        research methodology (literature review, experimentation…).</span></div>
    <div aria-live="assertive" id="magicdomid36" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> </span></div>
    <div aria-live="assertive" id="magicdomid37" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> Candidates
        should have the following skills:</span></div>
    <div aria-live="assertive" id="magicdomid38" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> •
        good proficiency in English, both spoken and written;</span></div>
    <div aria-live="assertive" id="magicdomid39" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> •
        scientific writing;</span></div>
    <div aria-live="assertive" id="magicdomid40" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> •
        programming (experience in C++ is a plus, but not mandatory).</span></div>
    <div aria-live="assertive" id="magicdomid41" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> </span></div>
    <div aria-live="assertive" id="magicdomid42" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">The
        candidate will be funded for 3 years; he/she is expected to
        defend his/her thesis and graduate by the end of the contract.
        The monthly gross salary is around 2000€, including benefits
        (health insurance, retirement fund, and paid vacations).</span></div>
    <div aria-live="assertive" id="magicdomid43" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> </span></div>
    <div aria-live="assertive" id="magicdomid44" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">For
        application, please send the following information in a single
        PDF file to Pr. Marius Bilasco (<a class="moz-txt-link-abbreviated" href="mailto:marius.bilasco@univ-lille.fr">marius.bilasco@univ-lille.fr</a>)
        and Dr. Fabio Morbidi (<a class="moz-txt-link-abbreviated" href="mailto:fabio.morbidi@gmail.com">fabio.morbidi@gmail.com</a>)with subject
        [PhD_EventSpike]:</span></div>
    <div aria-live="assertive" id="magicdomid45" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> *
        A cover letter.</span></div>
    <div aria-live="assertive" id="magicdomid46" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> *
        A curriculum vitae, including a list of publications, if any.</span></div>
    <div aria-live="assertive" id="magicdomid47" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> *
        Transcripts of grades of Master's degree.</span></div>
    <div aria-live="assertive" id="magicdomid48" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> *
        The contact information of two references (and any letters if
        available).</span></div>
    <div aria-live="assertive" id="magicdomid49" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"> </span></div>
    <div aria-live="assertive" id="magicdomid50" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">We
        look forward to receiving your application until </span><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g b"><b>12.04.2024</b></span></div>
    <div aria-live="assertive" id="magicdomid51" class="ace-line"
      spellcheck="true"><br>
    </div>
    <div aria-live="assertive" id="magicdomid52" class="ace-line"
      spellcheck="true"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g b"><b>Context
          of the joint PhD</b></span></div>
    <div aria-live="assertive" id="magicdomid53" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">The
        FOX research group is part of the CRIStAL laboratory (University
        of Lille, CNRS), located in Lille, France. We focus on video
        analysis for human behavior understanding. Specifically, we
        develop spatio-temporal models of motions for tasks such as
        abnormal event detection, emotion recognition, and face
        alignment. We are also involved in IRCICA (CNRS), a research
        institute promoting multidisciplanary research. At IRCICA, we
        collaborate with computer scientists and experts in electronics
        engineering to create new models of neural networks that can be
        implemented on low-power hardware architectures. Recently, we
        designed state-of-the-art models for image recognition with
        single and multi-layer unsupervised spiking neural networks. We
        were among the first to succesfully apply unsupervised SNNs on
        modern datasets of computer vision. We also developed our own
        SNN simulator to support experiments with SNN on computer vision
        problems.</span></div>
    <div aria-live="assertive" id="magicdomid54" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">Our
        work is published in major journals (Pattern Recognition, IEEE
        Trans. on Affective Computing) and conferences (WACV, IJCNN) in
        the field.</span></div>
    <div aria-live="assertive" id="magicdomid55" class="ace-line"
      spellcheck="true"><br>
    </div>
    <div aria-live="assertive" id="magicdomid56" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">The
        PR (Robotic Perception) team has skills in mobile robotics
        (perception), 3D reconstruction and unconventional vision. The
        PR team is piloting the e-Cathedral program (2015-2025) and is
        currently involved in four other projects which focus on event
        vision: the EVENTO project (2021-2024) co-financed by AID, the
        ANR CERBERE project ( 2022-2025) carried out by the LITIS
        laboratory of the University of Rouen, the ANR DEVIN project
        (2024-2028) carried by the I3S laboratory of the University of
        Côte d'Azur and the ANR-FWF project (France-Austria) (2024-2028)
        carried by the UPJV and by TU Graz. The PR team wishes to
        further strengthen this area of research and improve its
        expertise in AI through collaboration with the CRIStAL
        laboratory.</span></div>
    <div aria-live="assertive" id="magicdomid57" class="ace-line"
      spellcheck="true"><br>
    </div>
    <div aria-live="assertive" id="magicdomid58" class="ace-line"
      spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">The
        position is located in Lille, France. With over 110 000
        students, the metropolitan area of Lille is one France's top
        education student cities. The European Doctoral College Lille
        Nord-Pas de Calais is headquartered in Lille Metropole and
        includes 3,000 PhD Doctorate students supported by university
        research laboratories. Lille has a convenient location in the
        European high-speed rail network. It lies on the Eurostar line
        to London (1:20 hour journey). The French TGV network also puts
        it only 1 hour from Paris, 35 mn from Brussels, and a short
        trips to other major centres in France such as Paris, Marseille
        and Lyon.</span></div>
    <div aria-live="assertive" class="ace-line" spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"><br>
      </span></div>
    <div aria-live="assertive" class="ace-line" spellcheck="true">
      <div aria-live="assertive" id="magicdomid62" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">==</span></div>
      <div aria-live="assertive" id="magicdomid63" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">Ioan
          Marius BILASCO </span></div>
      <div aria-live="assertive" id="magicdomid64" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">Professeur
          des Universités en Informatique</span></div>
      <div aria-live="assertive" id="magicdomid65" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">Univ.
          Lille</span></div>
      <div aria-live="assertive" id="magicdomid66" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">CRIStAL
          CNRS - UMR 9189</span></div>
      <div aria-live="assertive" id="magicdomid67" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">Cité
          Scientifique</span></div>
      <div aria-live="assertive" id="magicdomid68" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g">59650
          Villeneuve d'Ascq Cedex - France   </span></div>
      <div aria-live="assertive" id="magicdomid69" class="ace-line"><span
class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g url"><a
            href="https://pro.univ-lille.fr/marius-bilasco/"
            rel="noreferrer noopener" class="moz-txt-link-freetext">https://pro.univ-lille.fr/marius-bilasco/</a></span></div>
    </div>
    <div aria-live="assertive" class="ace-line" spellcheck="true"><span
        class="author-a-hgz69zz77zz87zz77zz76zz82zoz83zz82zvz83zz68z1g"><br>
      </span></div>
  </body>
</html>