Efficient Deep Learning

Improving data training and computational efficiency for deep learning

Machine learning and in particular deep learning has dramatically improved the state-of-the-art in object detection, speech recognition, robotics, and many other domains. Whether it is superhuman performance in object recognition or beating human players in Go, the astonishing success of deep learning is achieved by deep neural networks trained with huge amounts of training examples and massive computing resources. Although already applied successfully in academic use-cases and several consumer products (e.g. machine translation), these data and computing requirements pose challenges for further market penetration.

EDL will significantly improve the applicability of deep learning, by creating data efficient training methods, and tremendously improving computational efficiency, both for training and inference. This requires a comprehensive approach that combines the domains of machine learning and computer systems: both are strong in the Netherlands but hardly connected.

EDL provides necessary innovations to improve efficiency in all areas, including simulated data, active learning, embedding model knowledge, visualization, platform mapping, low-power accelerators, data reduction, brain inspired spiking, and virtualization.

EDL solutions are widely applicable. EDL provides the vital steps to enable deep learning in the roadmaps of Dutch and international companies involved in this proposal; its need is clearly visible from the tremendous industrial interest.

Unique selling points:

  • Focused program, giving necessary efficiency improvements.
  • Holistic approach, attacking all algorithmic levels with computational co-design.
  • Synergy between machine learning and computational domains, both strong in the Netherlands, but hardly connected.
  • Coherent projects, interacting through challenging research lines.
  • Strong validation, many real-world demonstrators.
  • Huge impact, EDL techniques revolutionize industrial, societal, and economic systems.

EDL has received funding from NWO, the Dutch Organization for Scientific Research which falls under the responsibility of the Ministry of Education, Culture and Science. Next to the support of NWO-TTW, EDL is also financially and in-kind supported by 35 Dutch companies. This scientific research program is a cooperation of 7 Dutch Universities, 1 German University and 5 NWO research institutes.

Participating organisations

ASTRON
Eindhoven University of Technology
Natural Sciences & Engineering
Natural Sciences & Engineering
Netherlands eScience Center
Qualcomm (United States)
TATA Steel
Thermo Fisher Scientific (Netherlands)
University of Amsterdam

Impact

Output

Team

Rob van Nieuwpoort
Project Leader P4, member management board, Supervisor
University of Amsterdam
AB
Albert-Jan Boonstra
Co- Supervisor
Netherlands Institute for Radio Astronomy
Elena Ranguelova
Elena Ranguelova
Co-supervisor
Netherlands eScience Center

Related projects

Beyond the Data Explosion (eAstronomy)

An eScience infrastructure for huge interferometric datasets

Updated 20 months ago
Finished