Solve the social problems using Python

Nobukazu Ishii (~nobnov)


We have developed a system in which AI learns the captured video and predicts and notifies the action. Based on the captured data, it is judged whether the target person's behavior is normal or abnormal, and if it is abnormal, the family and staff are notified.

In Japan, the declining birthrate and aging population have become a social issue. Elderly people have difficulty living without support, but in Japan the number of supporters is decreasing due to the declining birth rate. In 1950, 12 workers of working age supported one elderly person, but in 2016, 2.2 persons are supporting one elderly person. Therefore, mechanization and automation of the support are urgent issues. This talk focuses on developing behavior analysis cameras using Python and the hard things leading up to implementation.

Basic outline of the talk

  1. Self introduction [2 min]
  2. Explanation of social problems [4 min]
    • Social problems in Japan (decreasing birthrate and aging of the population).
    • The aging countries.
  3. Solution selection [8 - 10 min]
    • Find a solution.
    • Reason for choosing a camera.
  4. System description [7 - 9 min]
    • Module, library.
    • Hard things in development.
  5. Reason for choosing Python [2 min]
  6. Q&A [3 - 5 min]

Who is this talk for?

  • Who has problems to solve in Python.
  • Who is interested in Machine Learning.
  • Who is interested in social problems.


Basic scripting in Python

Machine Learning

Speaker Info:

I am CEO of FANTOM Inc. We are developing behavior analysis cameras and nursing record systems for nursing homes in Japan.

Speaker Links:

Twitter :

Blog :

Company website :

Id: 1404
Section: Data Science, Machine Learning and AI
Type: Talks
Target Audience: Intermediate
Last Updated: