Vous êtes ici : Accueil » Kiosque » Annonce

Identification

Identifiant: 
Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?

Annonce

19 mars 2018

PhD position: Extraction and Characterization of Spatio-Temporal Patterns in Videos (Université de Tours, LIFAT)


Catégorie : Doctorant


Title: Extraction and Characterization of Spatio-Temporal Patterns in Videos

Keywords: video analysis; spatio-temporal patterns; sequence analysis; data mining; pattern recognition; fluid movement analysis

Supervision: Nicolas Ragot, MCF, HDR , Donatello Conte, MCF, HDR , Dominique Li , Julien Mille

Funding: Academic Grant

Location: Laboratoire d'Informatique Fondamentale et Appliquée de Tours (LIFAT, EA 6300), Université de Tours, France

 

Subject

This thesis research work is about video analysis. More specifically, we are interested in identification and characterization of trajectories in videos. Contrary to common trajectories corresponding to the movement of identifiable objects in videos (tracking of a person, a vehicle, etc.), the goal of this work will be rather on the analysis of the movement of a set of objects forming an heterogeneous "mass", like some fluids, human crowds, or animal swarms, for example. Therefore, with respect to the above context, this thesis will focus on three highly interdependent aspects:

The extraction of spatio-temporels patterns should allow to characterize locally the movements observable in the video ([7, 8]). In order to do this, we should be able to build some spatio-temporal descriptors based on dictionaries for example. One way to extract these patterns could be to use deep learning techniques. Once obtained, these patterns can be mixed together to build sequences.

The next step is to analyze these sequences to find more complex behaviours such as reoccurrence of subsequences, abnormal subsequences, etc. For this, we will focus on the use of data mining methods (sequence mining [1, 2, 3]), as well as sequence matching techniques [4, 5] and pattern recognition techniques [6], that will have to be adapted to the specificities of the videos. In particular, the goal will not be to work on a unique sequence but jointly with several ones that will characterize the behaviour of the heterogenous “mass” observed in the video.

The results of this thesis will be applied first to the analysis of fluid movements in engine cooling galleries, a problem related to an in-progress collaborative research work with researchers at Zhejiang University in China. Depending on results this shall lead to the submission of an international collaborative project to continue and expand the work. This work can also be applied to many other application domains, such as the analysis of heterogeneous fluid videos, the video-surveillance of human/animal populations, and, in particular, the analysis of crowds, especially on touristic sites.

Supervision

This Ph.D. thesis will be supervised by both BdTln and RFAI research teams of the LIFAT laboratory of the univsersité de Tours, France.

Directors: Nicolas Ragot, MCF, HDR - Donatello Conte, MCF, HDR

Supervisors: Dominique Li, Julien Mille

Requirements

Description of the position

Application details

The candidate should send by e-mail before the 21st of May 2018: a motivation letter, a detailed CV, the Master thesis abstract and scores obtained during master, the coordinates (e-mail and tel.) of one or two reference persons.

Host institution and place

The LIFAT Laboratory is composed of 47 faculty members including Professors, Associate Professors, Research Fellows and PhD students. The Laboratory is organized in three research groups involved on specific topics: Databases and Natural Language Processing, Scheduling and Control, Pattern recognition and Image analysis.

The scientific challenges addressed at the LIFAT Laboratory include the design of models, methods and algorithms to extract information, to discover knowledge from data, to solve combinatorial optimization problems by mainstreaming of human-computer interaction and considering optimization issues.

The Tours city is rich with history and has a well preserved heritage. The urban area of Tours (of nearly 300 000 inhabitants) has a leading part to play in the Loire Valley. It lies at the crossroads of the North-South and East-West communication lines of Europe and is only one hour from Paris by high-speed train. Cost of living is attractive.

References

  1. S. Ando and E. Suzuki. Discriminative learning on exemplary patterns of sequential numerical data. In ICDM 2014, pages 1-10, 2014.
  2. A. Bagnall, J. Lines, A. Bostrom, J. Large, and E. Keogh. The great time series classification bake off: A review and experimental evaluation of recent algorithmic advances. Data Mining and Knowledge Discovery, 31:606-660, 2017.
  3. J. Pei, J. Han, B. Mortazavi-Asl, H. Pinto, Q. Chen, U. Dayal, and M. Hsu. PrefixSpan: Mining sequential patterns by prefix-projected growth. In ICDE 2001, pages 215-224, 2001.
  4. Tanmoy Mondal, Nicolas Ragot, Jean-Yves Ramel, Umapada Pal, "Comparative Study of Conventional Time Series Matching Techniques for Word Spotting", Pattern Recognition, Vol. 73, pp. 47-64, 2018
  5. Tanmoy Mondal, Nicolas Ragot, Jean-Yves Ramel, Umapada Pal, Flexible Sequence Matching Technique: An Effective Learning-free Approach For word-spotting, Pattern Recognition, 2016, doi:10.1016/j.patcog.2016.05.011
  6. Yann LeCun, Yoshua Bengio. Convolutional networks for images, speech, and time series. The handbook of brain theory and neural networks. Pages 255-258. 1998
  7. Uijlings, J., Duta, I. C., Sangineto, E., & Sebe, N. (2015). Video classification with densely extracted hog/hof/mbh features: an evaluation of the accuracy/computational efficiency trade-off. International Journal of Multimedia Information Retrieval, 4(1), 33-44.
  8. Theriault, C., Thome, N., & Cord, M. (2013). Dynamic scene classification: Learning motion descriptors with slow features analysis. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 2603-2610).

Dans cette rubrique

(c) GdR 720 ISIS - CNRS - 2011-2018.