[visionlist] PhD position in lifelong visual SLAM - Bosch Corporate Research and University of Zaragoza

[Javier Civera]

------------------------------------------------------------------------------------------
1 PhD position available - Lifelong Visual SLAM
------------------------------------------------------------------------------------------

We have an unique opportunity to pursue doctoral research on visual SLAM, a
relevant topic of academic and industrial interest; and acquire an
excellent academic and industrial background from top actors in the field.
The focus will be lifelong visual SLAM, with the aim of developing scene
models that integrate the dynamics of typical scenarios, from visual data.

The PhD will be developed

- partly in the SLAMLab (University of Zaragoza, Spain), pioneers and
recognized experts in SLAM, authors of several of the most relevant SLAM
papers, under my supervision

- and partly in Bosch Corporate Research (Hildesheim, Germany), in one of
the biggest and most active computer vision groups in the European
industrial sector.

Please, see more information and apply in the following link:

https://www.bosch-career.de/en/bewerben/jobsearch/-/cui/
job/ZRB_UNREG_SEARCH/en/567D863A01021ED7B58752DFFC2C4786

------------------------------------------------------------------------------------------
PhD topic summary
------------------------------------------------------------------------------------------

Visual SLAM (Simultaneous Localization And Mapping) is one of the key
technologies for robotics, autonomous driving, Augmented- and
Virtual-Reality. Visual SLAM takes the images from a camera mounted on a
moving platform, with the aim of estimating its ego-motion and a scene map
from the video sequence.

Lifelong SLAM refers to the development of map models that capture
long-term dynamics of the objects in natural scenes (e.g., changes in
position of books on a desk, doors closing and opening, chairs that are
moved). Most of the current visual SLAM algorithms assume a rigid scene and
cannot model such changes, limiting the temporal validity of the maps.
Thus, the core topic of the PhD constitutes of the incorporation of such
dynamic patterns to SLAM maps, which is one of the less explored research
problems in the field.

------------------------------------------------------------------------------------------
Job description
------------------------------------------------------------------------------------------

* Research in the field of Lifelong Visual SLAM

* Develop and evaluate novel algorithms for Visual SLAM

* Develop efficient Computer Vision algorithms and software

* Analysis of research community, cooperate with and contribute to the
research in the host labs

* Publish scientific articles with the developed algorithms and results

* Attend scientific conferences and events to present your research results

* Prepare a doctoral thesis related with the research project

------------------------------------------------------------------------------------------
Requirements
------------------------------------------------------------------------------------------

* Very good Master degree(s) in Engineering, Physics, Robotics, Computer
Science or a related field, with a solid background in machine learning
and/or computer vision

* Excellent academic background

* High personal initiative, independent and structured way of working

* Demonstrable programming skills, preferably C/C++, Matlab and/or Python

* Demonstrable practical experience and knowledge in Machine Learning,
Computer Vision and/or visual SLAM

* Convincing appearance, high commitment combined with interpersonal skills
in teams

* Fluent English, written and spoken, knowledge of academic English

* German knowledge is an advantage

-- 
Javier Civera
jcivera[at]unizar[dot]es
http://webdiis.unizar.es/~jcivera
(+34) 876555554 <+34%20876%2055%2055%2054>

Departamento de Informatica e Ingenieria de Sistemas
Universidad de Zaragoza
C/ Maria de Luna, 1
50018 Zaragoza, Spain.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://visionscience.com/pipermail/visionlist_visionscience.com/attachments/20180103/ef378e7c/attachment.html>