Scientific Calendar Event
GO TO SCIENTIFIC CALENDAR



  • Europe/Rome
  • English
  • Administration
Description
An ICTP Virtual Meeting

TinyML enables machine learning technologies to perform on-device analytics of sensor data at extremely low power. This allows for new scientific applications to be developed at an extremely low cost and at large scale.
 
In recent years, hardware advancements have made it possible for microcontrollers to perform calculations much faster. Improved hardware has made it easier for developers to build programs on these devices. Perhaps the most important trend for scientists has been the rise of tiny machine learning, or TinyML.
 
Between hardware advancements and the TinyML community’s recent innovations in machine learning, it is now possible to run increasingly complex deep learning models directly on microcontrollers. TinyML represents a collaborative effort between the embedded power systems and machine learning communities, which traditionally have operated independently.
 
Topics:
  • ML general concepts
  • Scientific Applications of ML
  • Introduction to TinyML
  • Examples of TinyML applications

Speakers:
B. PLANCHER, Harvard University, USA
V.J. REDDI, Harvard University, USA
M. ROVAI, UNIFEI, Brazil
S. STINCKWICH, United Nations University Institute in Macau, China

Material

SciTinyML: Scientific Use of Machine Learning on Low-Power Devices | (smr 3612)
Go to day
  • Monday, 18 October 2021
    • 15:00 - 17:50 Introduction to Embedded ML (tinyML) 
      Time zone: You can choose your time zone (in place of Europe/Rome) on the top right of the page and the programme time will be modified accordingly.

The available slides and videos are published on SciTinyML web site
      • 15:00 Conference Intro: Why TinyML and Science 15'
        Speaker: Marco ZENNARO (ICTP, Italy)
      • 15:15 Intro to TinyML 30'
        Speaker: Vijay Janapa REDDI (Harvard University, USA)
      • 15:45 Q&A 30'
      • 16:15 Break 15'
      • 16:30 Intro to ML 45'
        Speaker: Laurence MORONEY (Google Inc., USA)
      • 17:15 Q&A 20'
      • 17:35 Outline for Full Conference 15'
        Speaker: Marco ZENNARO (ICTP, Italy), Brian PLANCHER (Harvard University, USA)
  • Tuesday, 19 October 2021
    • 15:00 - 17:40 
      The available slides and videos are published on SciTinyML web site
      • 15:00 Day Intro 15'
        Speaker: Marco ZENNARO (ICTP, Italy), Brian PLANCHER (Harvard University, USA)
      • 15:15 Hands on TinyML with KWS 2h0'
        Speaker: Brian PLANCHER (Harvard University, USA)
        Material: Slides
      • 17:15 Mark Multi-lingual KWS & Q&A 20'
        Speaker: Mark MUZUMDER (Harvard University, USA)
      • 17:35 Day Closing 5'
  • Wednesday, 20 October 2021
    • 15:00 - 17:55 Edge Impulse 
      The available slides and videos are published on SciTinyML web site
      • 15:00 Day Intro 15'
        Speaker: Marco ZENNARO (ICTP, Italy), Brian PLANCHER (Harvard University, USA)
      • 15:15 Intro to Iot/Sensors 30'
        Speaker: Sebastian BUETTRICH (IT University of Copenhagen, Denmark)
      • 15:45 Q&A 15'
      • 16:00 Break 15'
      • 16:15 Ethical Issues in ML and IoT 45'
        Speakers: Serge STINCKWICH (United Nations University Institute in Macau, China), Attlee GAMUNDANI (United Nations University Institute in Macau, China)
        Material: Slides
      • 17:00 Q&A 15'
      • 17:15 Breakout Discussion 20'
      • 17:35 Larger Group Discussion 15'
      • 17:50 Day Closing 5'
        Speaker: Marco ZENNARO (ICTP, Italy), Brian PLANCHER (Harvard University, USA)
  • Thursday, 21 October 2021
    • 15:00 - 18:00 
      The available slides and videos are published on SciTinyML web site
      • 15:00 Day Intro 15'
        Speaker: Marco ZENNARO (ICTP, Italy)
      • 15:15 Hands on  TinyML with Anomaly Detection 2h0'
        Speaker: Marcelo ROVAI (UNIFEI, Brazil)
        Material: Slides
      • 17:15 Break 10'
      • 17:25 Scientific Applications of ML 20'
        Speaker: Matthew STEWART (Harvard University, USA)
      • 17:45 Q&A 10'
      • 17:55 Day Closing 5'
        Speaker: Marco ZENNARO (ICTP, Italy)
  • Friday, 22 October 2021
    • 15:00 - 18:00 
      The available slides and videos are published on SciTinyML web site
      • 15:00 Day Opening and Where to go from here 20'
        Speaker: Marco ZENNARO (ICTP, Italy)
      • 15:20 Q&A 15'
      • 15:35 AI and EDU 20'
        Speaker: Hal SPEED
      • 15:55 Q&A 15'
      • 16:10 Break 10'
      • 16:20 Responsible AI / TinyML 30'
        Speaker: Susan KENNEDY (Boston University, USA)
      • 16:50 Q&A 15'
      • 17:05 Closing Keynote 30'
        Speaker: Pete WARDEN (Google Inc., USA)
      • 17:35 Q&A 15'
      • 17:50 Workshop Closing 10'
        Speaker: Marco ZENNARO (ICTP, Italy)
GO TO SCIENTIFIC CALENDAR