Skip to main content

KCGI Prof. Mochizuki and Ms. Sun's presentation in AI field won the best domestic award!

On November 11, 2020, Dr. Badru Mochizuki, Assistant Professor at Kyoto College of Computer Science and Graduate Institute of Information (KCGI), and Dr. Yimeng Sun's team presented their work at the ITU AI/ML in 5G Challenge, which aims to solve real-world problems by applying machine learning to communication networks, and were selected as one of the top three teams in Japan and won the Grand Prize.Dr. Mochizuki and Dr. Sun will participate in the ITU Final Conference, which will be held online worldwide in December.

Dr. Mochizuki and Dr. Sun selected "Network State Estimation by Video Analysis in Real-Time Streaming Service" from the three themes presented by the conference, and estimated the network state (throughput and loss rate) by using machine learning in real time.

This theme is a timely research field that can contribute to solving problems unique to modern society, where the worldwide spread of a new coronavirus has led to a rapid increase in the use of Zoom and other webcam-based telework systems, resulting in extreme congestion.The machine learning model developed by the team achieved a recognition accuracy of 96.3% as a network state estimation using deep learning from a given dataset of real-time videos.

The model constructed by Dr. Mochizuki and Mr. Sun is based on a neural network, and

  1. (1) Lightweight model: low computation time (can estimate the network state of 500 videos in 1 second)
  2. Portable: Platform-independent

These advantages led to the awarding of the Grand Prize.

Machine learning now influences the nature of communication networks, which are the lifeline of society, and this presentation proposes one of the ways in which machine learning can be applied to future networks.This model has the potential to improve communication speed, communication delay, and the communication system itself, and is expected to be used in the future.