ICML 2016 Workshop on On-Device Intelligence


Abstract


Consumer adoption of mobile devices has created a new normal in computing: there are now more mobile devices on the planet than people, and exabytes of mobile data per month now dominates the global internet traffic. As computing systems, these pocket-sized devices are more powerful in many ways than vintage supercomputers. They come packed with an ever growing array of sensors. They are “always-on”, and becoming increasingly capable of rich contextual understanding and natural interaction with their users.


This workshop will focus on research themes emerging at the intersection of machine learning and mobile systems. The topics of interest range from the design of new machine learning algorithms under storage and power constraints,  new on-device learning mechanisms, the interaction between devices and cloud resources for privacy-aware distributed training, and opportunities for machine learning in the nascent area of “Internet of Things.” The scope of the workshop also extends to real-time learning and optimization in the context of novel form-factors: wearable computers, home intelligence devices, and consumer robotics systems. We are also interested in hardware-software co-design for mobile machine learning applications.


Topics of interest include:

  • On-device inference

    • Techniques for model compression and quantization
    • Modeling innovations designed for size and speed
    • Power-aware energy-efficient models.
  • On-device learning 

    • Real-time Optimization
    • Continuous Learning: One-shot Learning, Transfer learning, Online unsupervised learning
    • Personalization and Adaptation
    • Devices and the Cloud

      • Distributed training on multiple devices
      • Privacy-preserving techniques for data collection and model training
      • Device-cloud interactions
      • Real-time Applications

        • On-device Language and Perception: Text, Speech, Images, Video
        • Sensors and Geolocation
        • Mobile healthcare
        • Perception and Control: Mobile Robotics
        • Internet of Things
      • Novel Form-factors, including Wearable Computing, Home intelligence devices
      • Hardware-Software Co-design


      Impact and expected outcomes 
      Mobile hardware, data, and usage patterns present a fundamentally different set of tradeoffs to machine learning algorithm designers and open up new opportunities for machine learning applications. We expect this workshop to be a forum for researchers tackling mobile machine learning from a variety of perspectives. Because this forum is novel, we expect the exchange of ideas across application areas and techniques to generate new collaborations and advancements. We hope to foster a subcommunity of researchers working with on-device intelligence and to define new research directions for the machine learning community as a whole.

      Impact and expected outcomes 
      One-day workshop comprising of a series of two keynote presentations, several invited and contributed talks.
      Subpages (1): Accepted Papers
      Comments