Two PhD positions
Mälardalens Högskola / Högskolejobb / Västerås
Observera att sista ansökningsdag har passerat.
Visa alla högskolejobb i Västerås,
Hallstahammar,
Surahammar,
Eskilstuna,
Enköping eller i
hela Sverige Visa alla jobb hos Mälardalens Högskola i Västerås,
Eskilstuna,
Arboga,
Flen,
Södertälje eller i
hela Sverige At Mälardalen University people meet who want to develop themselves and the future. Our 16 000 students read courses and study programmes in Business, Health, Engineering and Education. We conduct research within all areas of education and have internationally outstanding research in future energy and embedded systems. Our close cooperation with the private and public sectors enables us at MDH to help people feel better and the earth to last longer. Mälardalen University is located on both sides of Lake Mälaren with campuses in Eskilstuna and Västerås.
At the School of Innovation, Design and Engineering our students are studying to be for example innovators, entrepreneurs, illustrators, communications officers, network technicians and engineers. Here we have the research specialisations of Embedded Systems, and Innovation and Product Realisation. Our work takes place in cooperation with and in strategic agreements with companies, organisations and public authorities in the region.
Employment information
Employment: Temporary employment
Scope: Full time
Closing date for application: 2020-06-26
Campus location: Västeras
School: School of Innovation, Design and Engineering, (IDT)
Position description
We are offering two PhD-student positions in design and optimization of robust deep neural networks (DNNs) for safety-critical embedded systems, within the AutoDeep project. This is a collaborative project between MDH, Volvo and Zenuity funded by VINNOVA.
The first position will focus on developing an automatic framework to design a highly optimized set of DNN architectures to be deployed on embedded computing platforms. Inspired by the adequacy of convolutional neural networks in visual feature extraction and the efficiency of recurrent neural networks in dealing with temporal dependencies, the aim is to design optimized hybrid DNN architectures in capturing dynamic temporal dependencies in the context of detecting intention.
The second position is on improving the robustness of DNNs for safety-critical applications considering the faults in network model, the fragility in the training data, and also considering the robustness against unintended data perturbations. The effect of the adversarial and unintended data perturbations on the robustness of deep neural networks is also considered to be investigated.
As a PhD student, you will spend a minimum of 80% of your time on research studies. The rest will be spent on educational and/or administrative duties. The temporary employment is valid for 4 years.
Qualifications
Only those who are or have been admitted to third-cycle courses and study programmes at a higher education may be appointed to doctoral studentships. For futher information see Chapter 5 of the Higher Education Ordinance (SFS 1993:100).
Required qualifications:
• Master of Science in Computer Science, Computer and electrical engineering or equivalent
• Fluent in English, both written and spoken;
• Documented experience with vision and deep neural networks is of high merit.
• Experience with heterogeneous System Architecture like GPU and FPGAs is important
Decisive importance is attached to personal suitability. We value the qualities that an even distribution of age and gender, as well as ethnic and cultural diversity, can contribute to the organization.
Application
Application is made online. Make your application by clicking the "Apply" button below.
The applicant is responsible for ensuring that the application is complete in accordance with the advertisement and will reach the University no later than closing date for application.
We look forward to receiving your application.
We decline all contact with recruiters and salespersons of advertisements. We have made our strategic choices for this recruitment.
Varaktighet, arbetstid
Heltid/ Ej specificerat
Publiceringsdatum2020-06-17Ersättningmånadslön
Så ansöker duSista dag att ansöka är 2020-06-26
Klicka på denna länk för att göra din ansökanFöretagMälardalens Högskola
Omfattning Detta är ett heltidsjobb.
Arbetsgivare Mälardalens Högskola (org.nr 202100-2916),
http://www.mdh.se Arbetsplats Mälardalens högskola
Jobbnummer 5268217
Observera att sista ansökningsdag har passerat.