Zero Latency Linear Video Coding

Web site of the ZL-LVC project

Internships and Ph.D.

UPDATE: ONLY POSITION 2 IS OPEN FOR APPLICATION!

 

Three positions are to be provided within the Zero-Latency Linear Video Coding (ZL-LVC) project https://zllvc.wp.imt.fr/, a project funded by the French National Agency for Research (ANR) for the 2020 Call for Projects. The positions are open at three of the four members of the project: LTCI/Telecom-Paris (Palaiseau), L2S/CentraleSupélec (Gif-sur-Yvette), and IEMN/UPHF (Valenciennes), the fourth being EKTACOM (Les Ulis). Another position at EKTACOM will be open during 2021. Each position will benefit from double advisors and collaboration with all the members of the project. 

Scientific context: the Zero-Latency Linear Video Coding (ZL-LVC) project

Exchange of information with extremely low latency is of paramount importance for many emerging applications, such as remote driving, drone control, or remote surgery operation. Nevertheless, video coding and network transmission delays are incompressible. 

Linear video coding (LVC) is an emerging framework for wireless video transmission to multiple users. Relying on linear operators only, LVC provides each receiver with video quality commensurate to its channel quality. LVC suffers, however, from a significant coding latency.

The Zero-Latency Linear Video Coding (ZL-LVC) project aims at proposing a low-to-zero latency end-to-end video coding and transmission system. The goal is to optimize the rate-latency-distortion trade-off for traditional, and LVC systems, benefitting from the inherent robustness of LVC and from the quality enhancement that can be achieved by using deep and reinforcement learning methods. The potential of ZL-LVC will be assessed via a demonstrator using software-defined radio platforms.

Partnership

The four members of the consortium are LTCI, L2S, IEMN, and EKTACOM. 

The LTCI (Laboratoire de Traitement et Communication de l’Information) is a research lab of Telecom Paris / Institut Polytechnique de Paris. It addresses research issues in signal and image processing, digital communications, computer networks, and computer science. LTCI members involved in ZL-LVC are Marco Cagnazzo (full professor and project coordinator) and Attilio Fiandrotti (associate professor with a strong background in robust video coding and transmission), and a Ph.D. student to be hired (and to be co-supervised with L2S). 

The L2S (Laboratoire des Signaux et Systèmes, UMR 8506) is a common research unit between Univ Paris-Saclay, CNRS, and CentraleSupélec involving more than 200 people, with about 100 permanent researchers/professors. Areas of interest of the unit mainly concern the fundamental and applied aspects of mathematics at the origin of the development of signal and image processing, information theory, cryptology, and control theory. L2S members involved in ZL-LVC are Michel Kieffer (professor), Giuseppe Valenzise (CNRS researcher), one Ph.D. student to be hired (to be co-supervised with EKTACOM), and a Post-doc to be hired (to be co-supervised with IEMN). 

The IEMN (Institute of Electronics, Microelectronics, and Nanotechnology, UMR 8520) is a research unit supported by the CNRS and three regional partners: the University of Lille, Université Polytechnique Hauts-de-France (UPHF, Valenciennes), and ISEN Yncréa (Lille). The Digital Communications group of IEMN has strong expertise in research fields including digital signal processing for communications, quality-oriented optimization of video delivery over error-prone networks, multi-carrier transmission, cognitive radio, and advanced UWB technology for intelligent transportation systems. IEMN members involved in the project are François-Xavier Coudoux (professor), Patrick Corlay (professor), Mohamed Gharbi (associate professor), Anthony Trioux (associate professor), Christophe Deknudt (part-time associate professor), in addition to one Ph.D. student co-supervised with LTCI.

EKTACOM has provided consulting and integration services for multimedia solutions since 2000. Its main mission is to design and build video solutions for its customers from the Broadcast, Institutional, Defence, or Corporate world.  EKTACOM has started to work on the optimization of the new MPEG video coding standard (VVC, Versatile Video Coding), and its members involved in the ZL-LVC project have relevant roles in MPEG standardization (D. Nicholson is Co-Head of the French Delegation).

Open positions

Position level

Three fixed-term, 100%-funded positions are available respectively at LTCI/Telecom-Paris (Palaiseau, Ile de France, France), L2S (Gif-sur-Yvette, Ile de France, France), IEMN (Valenciennes, Hauts de France, France). The three selected candidates will integrate the ZL-LVC project and become members of the consortium, benefitting from double advisors and the collaborations among the four research teams. 

For each of the three positions, two mutually exclusive levels of candidature are possible:

  1. Candidature to a Master degree internship to be followed by a Ph.D. program (open only to Master students)
  2. Direct candidature to a Ph.D. program (open only to graduated students)

In the case of candidature to the internship+Ph.D., if the student is selected, she or he takes an engagement of 40 to 42 months: 4 to 6 months of internship, followed by 36 months of the Ph.D. program. The internships start in Spring 2021 and finish in September 2021. The Ph.D. program will start right after. 

In the case of candidature for Ph.D., if the candidate is selected, she or he can start the Ph.D. as soon as possible. Technical times to set up the Ph.D. program are in the order of 6 to 8 weeks, thus a start in March 2021 is a reasonable guess. For candidates outside the EU, the process could take a few additional weeks. 

Scientific program and hosting team

Independently from the position level, the scientific program of each position is described in the following. 

  1. Deep learning/computer vision methods for low-latency video transmission. Hired by LTCI and co-supervised by L2S, the selected candidate will be affiliated to the Multimedia team of the LTCI. The position is located on the new Palaiseau campus of Telecom-Paris.  The scientific goal is the design of techniques based on computer vision methods (using neural networks and deep learning) for reducing the latency in video transmission systems. The position, which may be complemented with teaching activities, involves theoretical research, software implementation, and experimentation. The required skills are computer vision, video processing, deep learning, generative models. Additional skills include video compression and transmission.
  2. Systems for zero-latency video transmission. Hired by L2S and co-supervised by EKTACOM. The position is located on the Paris-Saclay campus (Gif-sur-Yvette).  The student will join the telecommunication and networking cluster of the L2S. This program is devoted to the development of ultra-low-latency video coding and transmission solutions. The position may be complemented with teaching activities, and it involves theoretical research and computer simulations. The candidate should have a strong background in signal processing for communication and / or in source coding. Some background knowledge of (deep) reinforcement learning techniques is also appreciated.  
  3. Hybrid schemes for zero-latency video transmission. Hired by IEMN / Université Polytechnique Hauts-de-France, and co-supervised by LTCI, the Ph.D. student will be affiliated to the Digital Communications group of the Institute of Electronics, Microelectronics, and Nanotechnology (IEMN), UMR-CNRS 8520. The position is located on the campus Mont Houy, Valenciennes. The successful applicant will become part of a team of 5 researchers working on wireless communications, digital video transmission, joint source-channel-protocol coding, and artificial intelligence for broadcast and ITS applications. The student will contribute to the design of the zero-latency hybrid video scheme. The position may be complemented with teaching activities. This position involves theoretical research, computer simulation, and software-defined radio implementation.

For information, a fourth position will be soon available. Hired by EKTACOM and co-supervised by L2S (CIFRE). The focus is on the zero-latency aspects of the demonstrator and rate-distortion-latency optimization. We expect to start this position in Fall 2021, possibly with a preparatory internship.

Application

The positions are to be filled as soon as possible and  could start in March 2021. The application to one or more positions must be done by filling the following on-line form: https://forms.gle/sYqkgVK28CVWoyqX9  

Any interested candidate should first prepare:

  • A Curriculum Vitae with full details about the academic courses and results (GPA, exams grades, diploma grades where available)
  • A cover letter or motivation letter explaining why you are interested in the position(s).

Then, the candidate should submit her or his candidature through the web form (preferred solution). The same information (Name, contacts, CV, and motivation letter) can be sent here by email. Only one application is needed, even in the case of candidature to multiple positions. Candidates are invited to not send a second candidature by email if they used the on-line form. 

Qualifications and degrees

  1. Candidature to an Internship+Ph.D. Only Master students (M2 level) can apply. The student must be pursuing a degree in one of the following areas: Signal Processing, Digital Communications, Wireless Communications, Machine Learning/AI, Computer Science, Image and Video Coding, Telecommunication Engineering.
  2. Direct candidature to a Ph.D. program. Only graduated students can apply. A master-level university degree in one of the following areas is required: Signal Processing, Digital Communications, Wireless Communications, Machine Learning/AI, Computer Science, Image and Video Coding, Telecommunication Engineering.

Requirements

Independently from the level of the application, the candidate is expected to have:

  • Very good proficiency in English language (oral and written)
  • Very good proficiency in at least one programming language among C/C++, Python, Matlab
  • Very good academic results

Personal qualities

The candidates are expected to have the following qualities: 

  • Enthusiasm for research
  • Problem-solving skills
  • Conscientiousness
  • Team-working skills
  • Critical and scientific thinking

Contacts