Qualcomm Technologies has laid out its vision for ubiquitous on-device artificial intelligence (AI) complementing cloud AI. In a press release the company said it "envision a world where AI makes devices, machines, automobiles, and things much more intelligent, simplifying and enriching our daily lives."
In 2007, Qualcomm started exploring spiking neuron approaches to machine learning for computer vision and motion control applications, and later expanded the scope of the research to look not just at biologically inspired approaches but artificial neural networks - primarily deep learning.
Fast forward to today, and Qualcomm has announced the acquisition of Scyfer B.V., a company affiliated with University of Amsterdam and focused on cutting-edge machine learning techniques. Scyfer has built AI solutions for companies worldwide and in a number of different industries, such as manufacturing, healthcare and finance.
"We started fundamental research a decade ago, and our current products now support many AI use cases from computer vision and natural language processing to malware detection on a variety of devices - such as smartphones and cars - and we are researching broader topics, such as AI for wireless connectivity, power management and photography," said Matt Grob, executive vice president, technology, Qualcomm Incorporated.
Many companies focus on the execution of AI workloads in the cloud, but Qualcomm is focused on the implementation of AI on end devices - smartphones, cars, robotics, and the like - to ensure that processing can be done with or without a network or Wi-Fi connection, the company said. The benefits of on-device AI include immediate response, enhanced reliability, increased privacy protection, and efficient use of network bandwidth.
Qualcomm Technologies continues to push AI research forward and is bringing cutting-edge machine learning technologies to the forefront. Examples of such efforts include:
- Advances in neural network techniques for semi-supervised and unsupervised training like generative adversarial networks (GANs), distributed learning, and privacy protecting
- Network optimization for on-device applications including compression, inter-layer optimizations, optimizations for sparsity, and other techniques to take better advantage of memory and space/time complexity
- And specialized hardware architectures designed to accelerate machine learning workloads with greater performance and energy efficiency in embedded devices
The acquisition of Scyfer brings with it a founder and renowned professor at the University of Amsterdam, Dr. Max Welling, which will help to further advance AI research and development at Qualcomm Technologies. Dr. Welling will continue his role as a professor at the University of Amsterdam, and the rest of the Scyfer team will continue to be based in Amsterdam.
In 2015, Qualcomm Technologies and the University of Amsterdam also established QUVA, a joint research lab focused on advancing the cutting-edge machine learning techniques for mobile and computer vision. Qualcomm said it’s excited to continue to work with the University of Amsterdam going forward, underscoring the importance of educating future generations in this important field.