PhD candidate in Deep Efficient Temporal Learning

Faculty of Science – Informatics Institute

Publication date
23 July 2018
Level of education
Master's degree
Salary indication
€2,266 to €2,897 gross per month
Closing date
7 September 2018
Hours
38 hours per week
Vacancy number
18-451

The Informatics Institute of the Faculty of Science is inviting applications for a PhD candidate to work on the research of Deep Efficient Temporal Learning. The successful candidate will be based in the QUVA Lab in the Machine Learning Lab (AMLab) of the University of Amsterdam (UvA) led by Prof. Arnold W.M. Smeulders, Prof. Cees G.M. Snoek, Prof. Max Welling and Dr Efstratios Gavves. QUVA Lab hosts 12 PhD candidates and postdoctoral researchers, and is a joint collaboration between the University of Amsterdam and Qualcomm. Qualcomm is the world leader in mobile chip design and AI on mobile devices. The UvA is a leading university consistently ranked in the top 50 worldwide, with a world-leading Computer Science research department. The goal of QUVA Lab is cutting-edge research on Computer Vision and Machine Learning, marrying the best of the academic and industrial worlds. 

The Faculty of Science, which QUVA belongs to, holds a leading position internationally in its fields of research and participates in a large number of cooperative programs with universities, research institutes and businesses. The faculty has a student body of around 6,000 and 1,500 members of staff, spread over eight research institutes and a number of faculty-wide support services. A considerable part of the research is made possible by external funding from Dutch and international organizations and the private sector. The Faculty of Science offers thirteen Bachelor’s degree programs and eighteen Master’s degree programs in the fields of the exact sciences, computer science and information studies, and life and earth sciences.

World-class research groups directly involved in deep learning are AMLAB (machine learning led by Prof. M. Welling), ISIS (computer vision led by Prof. C. Snoek), and ILPS (Information Retrieval led by Prof. M. de Rijke). Examples of industry funded research labs involved in deep learning are Qualcomm-UVA (QUVA) Lab (12 PhDs/Postdocs), Bosch-UvA DELTA Lab (10 PhDs/Postdocs), Philips Lab (4 PhDs/Postdocs) and SAP-UvA Lab (3 PhDs/Postdocs). We also have ongoing collaborations with Microsoft Research (2PhDs).

Project description

The research topic of the position is 'Deep Efficient Temporal Learning'. The research will focus on the temporal Machine Learning and Deep Learning, with primary applications in Computer Vision. To date Deep Learning largely focuses on static and stationary inputs, mostly images or manually preprocessed videos. Yet, in reality inputs often come in streaming, noisy and high-dimensional sequences, characterized by non-stationary statistics and high redundancy. A typical example is streaming video in the wild, having frames that typically change slowly but continuously over time. The current solutions involve direct transfer to the temporal domain of Deep Learning and Machine Learning models originally designed for static and stationary inputs, such as single images. However, learning temporal representations on such sequences is shown to be challenging and unsatisfactory, with respect to both accuracy and efficiency. One could turn the weakness into strength, and exploit the slow nature of sequence data, for unsupervised and efficient temporal learning. Radical and novel learning theories and frameworks are needed for temporal representation learning.

The goal of this project is Temporal Representation Learning framework for Deep and Efficient Sequential Networks. The primary application is on videos and spatio-temporal representations, video-based recognition, video compression and spiking networks. That said, the research is expected to apply also to similar domains with strong temporal nature such as speech compression.

The research will be supervised by Dr Efstratios Gavves. Collaborations inside the lab with other professors and students are welcome. The research will also involve collaboration with Taco Cohen in Qualcomm Netherlands.

Requirements

The PhD candidate has:

  • a MSc degree (or equivalent) in either Artificial Intelligence, Physics, Electrical/Computer Engineering, Computer Science, Physics or abother related field;
  • a solid understanding of Machine Learning and Deep Learning;
  • excellent mathematical foundations. Special emphasis on statistics and probability theory, calculus and linear algebra;
  • excellent programming skills. Preferably Python, or other similar languages;
  • fluent communication and writing skills, cooperative spirit, excellent command of English.

Creativity and high motivation are greatly appreciated! A scientific track record is a strong advantage.

Further information

For further information about this vacancy:

Appointment

The appointment will be full-time (38 hours a week) for a period of four years (initial employment is 18 months). Periodic evaluations will be held after 9 and 14 months, and upon positive evaluation, the appointment will be extended to a total of 48 months. The appointment must lead to a dissertation (PhD thesis). An educational plan that includes attendance of courses, summer and/or winter schools, and national and international meetings will be drafted for the PhD candidate. The PhD candidate is also expected to assist in teaching of undergraduate students.

The salary is in accordance with the university regulations for academic personnel. The salary will range from €2,266 (first year) up to a maximum of €2,897 (last year) gross per month (scale P) based on a full-time appointment. There are also secondary benefits, such as 8% holiday allowance per year and an end of year allowance of 8.3%. The Collective Labour Agreement for Dutch Universities is applicable.

Starting date is flexible, preferably sometime in the last quarter of 2018.

Why apply?

  • Competitive pay and good benefits;
  • top-50 university worldwide;
  • an excellent Machine Learning, Deep Learning and Computer Vision Ecosystem, one of the best in the World. A vibrant community of top senior and junior researchers;
  • close collaboration with a leading industrial vendor of Machine Learning, Deep Learning and Computer Vision on mobile. Possibilities of internships;
  • interactive, open-minded and a very international city;
  • excellent research and computing facilities. Science Park is located near the city centre and close to a very friendly and fun part of the city;

English is the working language in the Informatics Institute. As in Amsterdam almost everybody speaks and understands English, candidates need not be afraid of the language barrier.

Job application

The UvA is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees. We value a spirit of enquiry and endurance, provide the space to keep asking questions and cherish a diverse atmosphere of curiosity and creativity.

You may only submit your application online using the link below. We will accept applications until 7 September 2018. To process your application immediately, please quote the vacancy number 18-451. There are no guarantees that late or incomplete applications will be considered.

Please do not send or cc your application to Dr Efstratios Gavves, Prof. A.W.M. Smeulders, Prof. Cees G.M. Snoek, Prof. Max Welling or Taco Cohen. Only applications via the online process will be considered.

What to include in the application?

  • A motivation letter explaining why you are the right candidate (max. 1 page);
  • as a separate section of the motivation letter, describe your vision on how to approach the problem of Deep Temporal Learning (max. 2 pages. Solid and creative ideas will be greatly appreciated!);
  • a curriculum vitae of max. 3 pages;
  • a copy of your MSc thesis. If your thesis is not in English, a translated summary (max. 4 pages);
  • a complete record of BSc and MSc courses, including grades;
  • a list of projects you have worked on, describing briefly your contributions (max. 2 pages). Your coursework and projects should motivate your competence in Machine Learning, Mathematics and Programming;
  • at least two academic references (industrial references are also appreciated). #LI-DNP

No agencies please

Published by  University of Amsterdam