Research Themes

Laboratory of Pattern Recognition and Machine Learning

Themes

Daily Life with Elves

How will we co-exist with increasingly intelligent robots and artificial intelligence? One answer is to use them as tools to enrich our lives. We call this kind of idea “Elven IT”, where we use IT to protect us while we enjoy life. There are three conditions for IT to be elves:

  1. AANormally the elves should not be visible.
  2. They should appear when necessary.
  3. They should provide the necessary services.

Such ideas have a long history: they originally appeared with the name of “smart homes”, before being known as “ambient intelligence” and appearing again in the present-day Internet of Things (IoT). In smart homes, the goal was to develop sensors to place throughout the home and use them to help with daily life. The target shifted to protection and services in ambient intelligence, while IoT focuses on new sensing technologies and connecting each individual device to the Internet. Our Elven IT is similar to ambient intelligence, with the major difference that the above conditions prevent the IT devices from interfering with daily life.

We created the first pressure-sensitive chair that aunthenticates the person sitting in it. We developed a system using ceiling-mounted infrared sensors that constantly monitors where a resident is located, while taking privacy concerns into account. We use this to identify when elderly residents living alone have fallen and analyze other aspects of their daily life. We also develop theory and software for high speed, high performance pattern recognition, needed to identify individuals. This includes the world’s fastest system for identifying nearest-neighbors when a (sufficiently small) error rate is allowed.
image14

The video below is a demo for fun. It was made by student volunteers in a few days in respond to professor’s request.


[Other studies in our lab] (Follow the lab’s page link PRML Lab for the detail)

Patten Recognition, Machine Learning, Data/Web Mining, Online learning, Bandit Problems, Sensor Applications, Image Processing, Behavior Analysis, Data Expression

Faculty

Mineichi Kudo
Mineichi Kudo
Position
Professor
Specialized field
Pattern Recognition Theory and Practice, Statistical Learning, Ambient Intelligence, Elves ICT
Hobby
Drinking, Tennis, Reading, Studying with drinking,

Favor yourself. The aim in undergraduate education is to obtain basic skills, so that the lectures are not easy to understand. You might feel uncomfortable about learning without specific goals. However, once you got them, they will be your strong tools for tackling your own problems. Sharpen your tools. It supports you in your life-time. Do not regret to use time for it. By the way, when you start your own study, what if you don't have any idea to do ? For doing an original study, you have to have your own ideas or viewpoints unlike the others. If not, what your past effort is for. That is what I meant in the beginning. Favor yourself because you are unique in viewpoint and thinking. Some bitter experiences would contribute to build your original personality. Let's do something interesting so as to hear "it's interesting."

Atsuyoshi Nakamura
Atsuyoshi Nakamura
Position
Associate Professor
Specialized field
Online learning, string algorithms, network analysis
Hobby
Jogging, music

I welcome your questions. I was also a question monster in my primary school days. Most students have not understood what you cannot have understood. I have learned many things from student questions. So, please ask questions not only for yourself but also for me.