PhD student in Computer Engineering for Vision Transformer and AoI

At the department, education is conducted in the subjects of computer engineering, electronics, electrical engineering and sound production. The research at the department takes place at the research center Sensible Things that Communicate (www.miun.se/stc) and deals with sensor-based systems and services for use within the Internet of Things and artificial intelligence.

The subject Computer engineering for third cycle education at Mid Sweden University consists of three specializations, Communication Networks, Data Analysis and Multidimensional Signal Processing. This doctoral student position has a more theoretical approach relating to the specializations through applications.

Job description

This research project seeks to innovate at the intersection of advanced vision transformer architectures and dynamic data processing. It aims to develop a versatile framework capable of adapting to various real-time applications, addressing the need for balanced support across different research branches and paving the way for breakthroughs in efficient data processing and analysis. Unlike existing works that focus on either improving the transformer architectures for specific domains or developing efficient and effective techniques for dynamic data processing, this research project aims to integrate these two aspects and develop a learning-based framework that can adapt to various real-time applications. The concept of Age of Information (AoI), which has gained significant attentions in networked systems, is incorporated to improve the timelessness of the information flow of the transformer architecture as well as to support real-time adaptive update architecture parameters, which combined address the challenges associated with the computational overhead of vision transformers in general.

In the project the doctoral student will be responsible for:

• Developing and implementing a transformer-based framework that is adaptable for diverse real-time data processing needs.
• Undertaking case studies with this framework to explore applications like light field video compression optimization and improving information freshness in dynamic learning environment.
• Performing validation studies to evaluate the framework's effectiveness in practical settings, including addressing any ethical concerns.
• Assessing the framework's potential to improve real-time data processing in both visual and non-visual contexts.

As a doctoral student you will conduct independent third-cycle studies and is expected to conduct both theoretical and experimental research work.


Entry requirements

General entry requirements include a second-cycle level qualification, or fulfilment of courses comprising at least 240 credits of which at least 60 credits must be at second-cycle level, or the equivalent knowledge gained in some other way in Sweden or abroad. Specific entry requirements include at least 90 credits in the subject of Computer Engineering, Computer Science, Electrical Engineering, Engineering Physics, Applied Mathematics, or the equivalent knowledge gained in some other way in Sweden or abroad.

Assessment criteria

We are looking for candidates with:

• a strong interest in learning-based data processing and its applications in networked systems and computer vision,
• a solid foundation in of mathematics, especially linear algebra
• proficiency in programming (preferably Python),

Experience and completed course in one or more of the following subjects are highly valued:

• machine and deep learning (theoretical and practical use)
• computer vision, image and video processing,
• communication and networked systems.

Knowledge about transformer architectures, optimization and information theory is a plus.

Personal qualities

As a doctoral student you should show:

• a high level of creativity, and innovative thinking,
• thoroughness and attention to research and development details,
• a systematic and structured approach to problem-solving and challenges,
• ability to collaborate effectively within team settings, as well as work independently,
• a strong ability to convey complex concepts and ideas clearly and effectively.

Other assessment criteria

Other skills relevant to the doctoral position is to exhibit:

• effective communication abilities, with a strong command of English in both verbal and written communications,
• comprehensive knowledge of standard software to support both research-related, teaching, and general administrative duties.

Employment process

Processing of the appointment will comply with the provisions in Chapter 5 of the Higher Education Ordinance, and will be carried out in accordance with Mid Sweden University’s Employment Procedures. 

Terms of employment

The doctoral studentship position corresponds to four years of full-time studies and is expected to lead to a Degree of Doctor. The employment may include up to 20% other duties such as teaching, in which case the duration can be extended to a maximum of 5 years. Start date 2024-09-01, or by agreement.

Initial doctoral student positions are valid for a maximum of one year, after which the position may be renewed by no longer than two years at a time. Provisions regarding doctoral studentship employment can be found in the Higher Education Ordinance (1993:100), Chapter 5, Sections 1-7.

Place of employment: Sundsvall

Salary: In accordance with Mid Sweden University’s salary scale for doctoral students.  

Information

For more detailed information contact:
Associate Professor, Stefan Forsström (Stefan.Forsstrom@miun.se), Dr. Qing He (qinghe@miun.se or +46 10 142 82 75), Dr. Roger Olsson (Roger.Olsson@miun.se or +46 10 142 86 98), or Head of department Associate Professor Patrik Österberg (Patrik.Osterberg@miun.se or +46 10 142 86 14).

Application

All application documents shall be submitted in Swedish or English, and must include:

1)  a personal letter (at most 2 pages) including a motivation for why you are applying for this position, and a self-assessment on why you are the right candidate for this position,
2)  a curriculum vitae (CV) with any experiences you find relevant for this position,
3)  degree diplomas (BSc and MSc) and transcript of records with signed grades, 
4)  the master thesis (or a draft thereof), and/or some other self-produced technical or scientific text), publications, and other relevant documents, 
5)  references with contact information (names, emails, and telephone number), and up to two letters of recommendation.

All applicants should state their earliest possible starting date.

Welcome with your application through our recruitment system no later than 2024-04-30.

Mid Sweden University has two campuses; in Sundsvall and in Östersund, both located close to students, colleagues and the surrounding society. At the same time, it is located conveniently close to the sea, forests and mountains, which provides a great quality of life. Here, new knowledge is created by means of internationally successful research and education that contribute to the development of our society. This way, we actively contribute to a sustainable future and a better world. Welcome to a university where people meet, get inspired and think in new ways.  Mid Sweden University works actively for equal opportunities and strives to embrace the qualities that diversity and equality bring to the organization. Prior to any recruitment work, Mid Sweden University has taken a position on consulting support, recruitment channels and marketing. We therefore kindly ask media vendors, recruitment sites and the like not to contact us.

• Contract type

  Full time

• First day of employment

  Start date 2024-09-01 or by agreement

• Salary

  Salary in accordance with Mid Sweden University’s salary scale for doctoral students

• Number of positions

  1

• Working hours

  100%

• City

  Sundsvall

• County

  Västernorrlands län

• Country

  Sweden

• Reference number

  MIUN 2024/536

• Union representative
  1. Per Bergman, Fackförbundet ST, 010-1428371
  2. Börje Norlin, Saco, 010-1428594
• Published

  2024-02-27

• Last application date

  2024-04-30