CYPRESS Video Analytic Framework

High definition acquisition devices have been commonly used in video capture. Millions of videos are recorded and stored daily. Due to the rapidly growing volume, effectively searching and matching the desired video clips from archives has become increasingly challenging. It may need days and weeks to locate the target information, such as suspected criminals, traffic violated vehicles, or accidental fall in care homes. While Google engine is designed for web searching, we propose to build a framework for efficient video searching. We will evaluate various approaches, in particular state-of-the-art machine learning techniques, parallel computing, real-time transmission and big data storage. By exploiting these technologies, we expect to deliver a real-time video searching and matching system, which is robust and can be deployed in a large application domain.

Faculty Supervisor:

Irene Cheng

Student:

Partner:

Istuary Innovation Labs Inc (Vancouver, BC);Eyexpo Technology Corp;eChaperone.AI

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Alberta

Program: