[visionlist] PhD position in continual deep learning for computer vision

[Jose Dolz]

Applications are invited for a fully funded PhD position at the ETS,
Montreal, Canada. ETS is the fastest-growing and largest engineering school
in Quebec, with an expanding team of highly qualified young researchers in
image analysis, computer vision and deep learning, some of the priority
areas of the school.

The position is available after the candidate passes ETS application
requirements and the candidate will start at her/his convenience (latest at
summer 2021). Financial support is available for 4 years. This project will
explore learning strategies for training under a sequential stream of
training data, to alleviate the potential problem of catastrophic
forgetting in deep neural networks. The main application domain will be in
computer vision. The successful candidate will work under the supervision
of Prof. Jose Dolz and will potentially collaborate with other Professors
and research groups. Furthermore, the selected student is expected to
publish her/his research on top computer vision journals (IJCV) and
conferences (CVPR, ECCV, ICCV, etc).

Prospective applicants should have:

   -

   Strong academic record with an excellent M.Sc. degree in computer
   science, applied mathematics, or electrical/biomedical engineering,
   preferably with expertise in more than one of the following areas: medical
   image analysis, machine learning, computer vision, pattern recognition,
   semi/weakly supervised learning and/or optimization;
   -

   Experience with a deep learning framework (preferrably PyTorch, or
   Tensorflow).
   -

   Knowledge of weakly or semi-supervised learning strategies is an asset.
   -

   Publications in a peer-reviewed journal or conference in a related topic
   are a bonus.


For consideration, please send a CV, a cover letter, names and contact
details of two references, transcripts for graduate studies, and a link to
a M.Sc. thesis (as well as relevant publications if any) to:

jose.dolz at etsmtl.ca

-- 
*Jose Dolz*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://visionscience.com/pipermail/visionlist_visionscience.com/attachments/20200630/f0ba780a/attachment.html>