Research for more efficient self-learning systems

NWO supports multidisciplinary collaboration, with a focus on use

04 DEC 2017

A computer which can autonomously recognise dangerous situations from security images: self-learning automated systems make this possible. A consortium of which SURFsara is a member has been awarded financing from NWO for a project to make deep learning - where computers learn new things independently - more efficient and transparent.

Self-learning systems

A computer which can autonomously recognise dangerous situations from security images: self-learning automated systems make this possible. But before these kinds of systems can operate independently, you have to first design it and then train it using a huge number of examples. The system also needs a lot of processing power to be able to make decisions. Even then, it's not always clear which specific elements the system has actually based its decision on.

Intelligent hearing aids

Researchers will make deep learning much more efficient and transparent within the Efficient Deep Learning programme, using examples from everyday life. They hope this will make the technology useful to automatic visual inspections, tissue analysis, smart equipment maintenance and intelligent hearing aids that work well in noisy environments, among other things. SURFsara is contributing to the programme by supplying expertise and hardware resources, and playing an active role in the user committee.

Project participants:

AIIR Inovations, ASTRON, CWI, Cyclomedia, Cygnify, Donders Instituut, FEI, 2getthere, GN Hearing, Holst Centre, ING, Intel, Irdeto, Lely, Mobiquity, NLeSC, NXP, NVIDIA, Océ, Radboudumc, Schiphol, Scyfer, Sectra, Semiotic Labs, Siemens, Sightcorp, Sorama, SURFsara, TASS International, Tata Steel, Technische Universität Dresden, Delft University of Technology, Eindhoven University of Technology, Thales, TNO, TomTom, University of Twente, University of Amsterdam, 3DUniversum, VicarVision, ViNotion, Vrije Universiteit Amsterdam and Wageningen University & Research.

Perspectief Financing Programme

The management of NWO's Applied and Technical Sciences section has made 21 million euro available for six comprehensive research programmes as part of the Perspectief financing programme. The companies, civil society organisations and knowledge institutes involved are supplementing NWO's investment to the tune of 11 million euro. The overall budget allows 74 doctoral candidates and 25 postdocs to work over the next 5 to 6 years.

Collaboration between industry and social organisations

With Perspectief, NWO, Netherlands Organisation for Scientific Research, is challenging scientists to establish close collaboration between industry and civil society organisations. Their research will be multidisciplinary, with a clear focus on use. Together, the parties will establish new research lines which tie in with the top sectors of industry.

More information:

Website NWO

 

Latest modifications 06 Dec 2017