AI-based Drowning Detection Using Surveillance Video Camera
Project Description
The World Health Organization estimates around 300,000 drowning deaths per year globally. The launch of Dolphin 3 rescue robot https://www.yunzhou-tech.com/info/363.html helps to save drowning people in open-water. However, the detection of drowning people is still not automized, thus affecting the efficiency of deployment of Dolphin 3. In this project, we aim to implement AI-based drowning detection methods to facilitate more efficient applications of Dolphin 3.
Supervisor
YU, Weichuan
Quota
5
Course type
UROP3200
Applicant's Roles
Train drowning detection methods using surveillance video data.
Applicant's Learning Objectives
• Learn basic knowledge of video analysis.
• Learn basic knowledge of deep learning methods.
• Train deep learning methods to optimize the drowning detection performance using surveillance videos.
• Test the performance of new methods using surveillance video data.
• Learn basic knowledge of deep learning methods.
• Train deep learning methods to optimize the drowning detection performance using surveillance videos.
• Test the performance of new methods using surveillance video data.
Complexity of the project
Challenging