Romeo Tatsambon Fomena, Web Page

| Contact | Vitae and position | Research | Publications | Demos | Teaching |

Contact at ECEmailto

Address: Department of Electrical & Computer Engineering
2nd floor ECERF
9107 - 116 Street
University of Alberta
Edmonton, Alberta, Canada T6G 2V4
Tel: (+1) 780 200 3209
Picture of Romeo Tatsambon Fomena

Contact at CSmailto

Address: Department of Computing Science
2-21 Athabasca Hall
University of Alberta
Edmonton, Alberta
Canada T6G 2E8

Vitae and position

I was born in 1982 in Cameroon, where I received a BSc in Mathematics and Computer Science in 2002 from the University of Ngaoundere.

On the completion of my bachelor degree in 2002, I received a grant from the French government and the Brittany council enabling me to do postgraduate studies at ISTIC, Universite de Rennes 1, where I received in September 2005 a MEng in Computer Science with major in Computer Vision, and I also received a MSc in Image Processing the same year from SPM, Universite de Rennes 1.

In 2005, I have spent six months at the University of Adelaide, Australia, for my end of study internship, working on coherence optimization in polarimetric SAR interferometry.

In Nov. 2005, I started a PhD degree in the Lagadic group of INRIA. My research, supervised by Francois Chaumette, was focused on vision-based control of robotic systems.

In August 2008, I was at the University of Illinois at Urbana-Champaign, USA, in the framework of a collaboration between Inria and the Beckman Institute. I have carried out experiments on visual servoing of a mobile robot using a catadioptric camera.

In September 2008, I have co-chaired the visual servoing session of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems, IROS'08.

In November 2008, I received my PhD degree from Universite de Rennes 1.

For the September 2008 to August 2009, I was a Teaching assistant at ISTIC, the Computer Science and Electrical Engineering department of Universite de Rennes 1.

From October 2009 to March 2011, I was a Postdoctoral researcher at IRSTEA, the French Environmental Sciences and Technologies Research Institute. The goal of my research, supervised by Christophe Collewet, was to investigate an active control of fluid flows using visual sensing.

Since May 2011, I am a Postdoctoral Fellow of the University of Alberta. The goal of my research, supervised by Alan Lynch and Martin Jagersand, is to investigate multiview geometry image-based task specification and control of a rigid body.

Research areas: control theory, computer vision, robotics, fluid mechanics and machine learning

Image-based control of a rigid body in SE(3)

My research focuses on developing theory for closed-loop vision-based control of dynamic system which includes fluid flows, manipulators, mobile and aerial robots.

Closed-loop vision-based control or visual servoing consists of using visual features extracted from a vision sensor to control the motion of a dynamic system. A vision sensor provides a large variety of possible visual features. From this large spectrum of potential visual features, a current opened research question is: how to select visual features corresponding to an ideal system behaviour?

Ideally, satisfaction of the following criteria is expected for the control law: local and as far as possible global stability, robustness to measurement and modeling errors, non-singularity, local minima avoidance, satisfactory motion of the robot and features in the image, and finally maximal decoupling and linear link (the ultimate goal) between the visual features and the degrees of freedom taken into account.

My PhD work has investigated the use of a spherical projection model to design decoupled control schemes in order to cope with robustness and stability issues of image-based visual servoing. Significant results have been obtained for static objects such as spherical objects, and volumetric and planar objects defined by points cloud. The main application can be found in the autonomous positioning of a robot (six DOFs manipulator or mobile robot) equipped either with a classical perspective camera or with an omnidirectional vision system.

Image-based control of fuild flow

My work at IRSTEA has focused on fluid flows control. Christophe Collewet and myself, have proposed for the first time a closed-loop vision-based approach for fluid flows control: see a detailed demo here vision-based control of Poiseuille flow. Fluid flows control has potential applications in sustainable development such as the efficiency improvement of wind turbines.

Image-based assistive teleoperation

In practice, human and robots collaborate together as a team to achieve a task. Now I am interested in reducing the fatigue of a human who is involved in the control loop. This research has applications in vision-based navigation and control of helicopter UAVs for inspection of power lines and, in vision-based telerobotics for service inspection and exploration tasks.

Professional activities

  • Reviewer for journal articles for Control Engineering Practice, the Journal of Intelligent and Robotic Systems, Robotics and Autonomous Systems
  • Reviewer for conference articles for IEEE RO-MAN'08, IEEE ICRA'10, IEEE/RSJ IROS'10, IEEE/RSJ IROS'11, IEEE ICRA'13, IEEE/RSJ IROS'13
  • Publications

    Complete list (with postscript or pdf files if available)

  • R. Tatsambon Fomena, C. Perez Quintero, M. Gridseth, M. Jagersand- Towards Practical Visual Servoin in Robotics.-
    Conf. on Computer and Robot Vision, May 2013. pdf

  • R. Tatsambon Fomena and C. Collewet- Fluid Flow Control: a Vision-based Approach.-
    International Journal of Flow Control, 3(2&3):133-169, September 2011.

  • R. Tatsambon Fomena, O. Tahri and F. Chaumette- Distance-based and Orientation-based Visual Servoing from Three Points.-
    IEEE Trans. on Robotics, 27(2):256-267, April 2011.

  • R. Tatsambon Fomena and F. Chaumette- Improvements on Visual Servoing from Spherical Targets using a Spherical Projection Model.-
    IEEE Trans. on Robotics, 25(4):874-886, August 2009.

  • R. Tatsambon Fomena and F. Chaumette- Visual Servoing from Spheres with Paracatadioptric Cameras.-
    Recent Progress in Robotics, Selected papers from ICAR'07, S. Lee, I. Hong Suh, M. Sang Kim (eds.), pp. 199-213, LNCIS 370, Springer-Verlag, 2008.

  • R. Tatsambon Fomena, H. Ul Yoon, A. Cherubini, F. Chaumette and S. Hutchinson- Coarsely Calibrated Visual Servoing of a Mobile Robot using a Catadioptric Vision Sensor.-
    IEEE/RSJ Int. Conf. on Intelligent Robots and Systems, IROS'09, St Louis, USA, October 2009.

  • R. Fomena and S. Cloude- On the Role of Coherence Optimization in Polarimetric SAR Interferometry.-
    Proceedings of International CEOS SAR CAL/VAL workshop, Adelaide, Australia, September 2005.
  • Demos

    Illustrations
    Visual interfaces for human-robot interaction: videoicone video
    Vision-based control of Plane Poiseuille flow: video icone video
    Visual servoing from a spherical ball with a paracatadioptric camera: videoicone video
    Visual servoing from a special compound of features: videoicone video
    Visual servoing from a special soccer ball with a fisheye camera: videoicone video
    Visual servoing of a mobile robot from a target point with a catadioptric camera: videoicone video

    Teaching

    Guest lecture, Department of Computing Science, University of Alberta

    Term: Winter 2012-2013

  • Introduction to Visual Servoing pdf
  • Towards Practical Visual Servoing in Robotics pdf
    Audience: 4th year Undergraduate students
  • Assistant Professor in Computer Science, Universite de Rennes 1

    During academic year 2008-2009, I was a assistant Professor at ISTIC, the Computer Science and Electrical Engineering department of Universite de Rennes 1.

    I was involved in various subjects in the field of computer science and computer vision.

  • 90h Introduction to programming in Java, Scheme and Pascal
    Audience: first year students

  • 48h Compilation
    Audience: master students
    ANTLR is the compiler-complier tool used to build a simple compiler for VSL+

  • 26h Graph theory and practice
    Audience: master students
    Java is the programming language used to implement graph algorithms

  • 110h Computer Vision, Image Processing, Stochastic Image Processing
    Audience: master students
    C++ is the main programming language used to implement camera calibration algorithms, image processing algorithms

  • 24h Introduction to Statistics
    Audience: master students
    SAS ACADEMIC is the main tool used to analyse statistic data

  • 12h Introduction to Computer Networks
    Audience: master students
  • logo of UA

    Copyright March 2013 © Romeo Tatsambon Fomena