Artificial intelligence is poised to fundamentally change our world: in future, machines and devices will be capable of autonomously learning from experience and acting on this basis. The foundation for this is deep learning. The University of Amsterdam (UvA) and Bosch will cooperate closely in this field by partnering in Delta Lab (Deep Learning Technologies Amsterdam). Ten PhD students and postdoctoral fellows will conduct research at Delta Lab, which will be located at Amsterdam Science Park.
The aim of this research alliance is to promote stronger exchange between fundamental and applied research through knowledge transfer, thereby generating new knowledge and innovations. The ten PhD students and postdoctoral fellows, led by UvA professors Max Welling and Arnold Smeulders, will work closely with researchers from the Bosch Center for Artificial Intelligence at Bosch’s Renningen research campus. In return, Bosch will send associates to work on joint scientific projects at the UvA.
‘Industrial basic research in the field of artificial intelligence benefits from close contact with academic institutions’, says Dr Michael Bolle, head of research and advance engineering at Bosch. ‘I’m looking forward to working with Professor Max Welling and his team from the University of Amsterdam. Worldwide, he is one of the leading names in the field of deep learning.’
‘Deep learning allows you to discover hidden structure in data and make predictions based on data’, says Max Welling. ‘The applications are numerous. For instance, think of fully autonomous cars, robots, smart homes and the internet of things.’
For their research projects, the PhD students and postdoctoral fellows will develop mathematical models and algorithms with which machines can learn from information and experience. ‘There is a great demand for smart devices and systems, which in turn only increases the importance of deep learning’, adds Welling. ‘Research findings from Delta Lab may result in applications and products in which Bosch is a global leader.
Arnold Smeulders adds: ‘Computer vision has the largest amount of sensory data. This is why this sub-field of artificial intelligence is one of the most exciting goals for deep learning and why the intersection of deep learning and computer vision is so important. It presents specific challenges to deep learning to learn from the enormous amount of data. There are so many areas in which the results can be applied: in intelligent cameras, self-driving cars, home care, etc. Actually, I wouldn’t know of areas where it could not be applied.’
The partnership with Bosch acts as a stimulus for research at the UvA’s Informatics Institute. This is the second public-private partnership the institute has entered into. In 2015, the QUVA lab was created, a partnership between the UvA and Qualcomm Technologies, Inc.