RoboCup Rescue Simulation with Deep Learning
Disasters such as earthquake and tsunami can cause significant destruction to a city and hurt many people. To reduce the amount of the dead troll, fast disaster response to rescue survivors in a disaster zone is of paramount importance. However, the problem is all the current method to manage disaster environment is all done by human and their work burden is too much to save the people as much as possible. Especially, search for the location of people who need rescue in the disaster zone spend a considerable amount of time, which is one of the most important issues in disaster relief. The one of the solution to solve this problem, if we can predict the location of injured people in a disaster situation, it can help the rescue team to deploy the rescue team more quickly and accurately so that the time to save people can significantly reduce. Therefore, in this study, we developed a software package for predicting the location of injured people in an earthquake situation based on deep learning. However, there are limitations of conducting an experiment in a real disaster situation and collecting real disaster data set to train the machine learning model. Therefore, we predict the hidden injured in the virtual disaster simulator called RoboCup Rescue Simulation (RCRS) with deep learning.
Leave a comment