Skip to content

Machine-Learning-Tokyo/Edge_AI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 

Repository files navigation

MLT Edge AI

We are excited to kick off our first Embedded ML workshop - a combination of presentation and ideathon. https://www.meetup.com/Machine-Learning-Tokyo/

Firstly, we will talk about the promise of light-weight deep neural networks for energy-efficient and low-cost IoT applications. We discuss some examples of accelerated and low-memory version of deep learning models for real-time use, predictive maintenance, time-series analysis, and demand forecast. We focus on AI methods for turning IoT data into insights and actions.

After the introductory talk about Edge AI we will build teams of 3-5 people and have 1.5 hours to come up with project ideas for embedded ML scenarios that include use case, feasibility assessment, workflow, allocation of resources (human, time, computation, ...).

-- AGENDA --

  • 12:30 pm Doors open
  • 1:00 pm - 1:45 pm Talk: "AI for Embedded Computing: Towards low-power Edge Inference", Hossein I. Rad
  • 1:45 pm - 2:00 pm Edge AI Ideathon, Yoovraj Shinde
  • 2:00 pm - 2:45 pm Team building break with snacks and drinks
  • 2:45 pm - 4:15 pm Ideathon
  • 4:15 pm - 5:00 pm Presentations
  • 5:00 pm Wrap up

-- MLT EDGE AI TEAM --

Hossein Izadi Rad is a Ph.D. candidate of Information Science and Technology with the University of Tokyo. He has an M.E. degree in Electrical Engineering. His recent works are concerned with off-the-cloud time-series prediction and demand forecast. He has collaborated with several tech startups in Tokyo.

Yoovraj Shinde is the Co-Founder of MLT and currently working as Technologist at Rakuten Institute of Technology. Electronics Engineer by heart, but worked in software industry for about 8 years. Worked on FrontEnd, BackEnd Java systems, iOS app development. His interests are Machine Learning and Robotics.

9 WORKSHOP TEAMS: IMPRESSIONS

Releases

No releases published

Packages

No packages published