Simultaneous Object Detection and Tracking
Object detection and tracking have lots of real world application e.g. surveillance, self driving cars and many more. Many off the shelf object detection algorithms are available that need need good GPUs for real time inference. Light weight object detectors are fast and can provide real time performance on edge devices and CPUs but are less accurate. Heavy weight object detectors are not fast but are more accurate, so their learning pattern can be distilled to light weight models. This project is all about developing a light-weight fast and accurate object detector and then tracker that can be easily deployed on edge devices i.e. AI Hubble cameras.
Keywords: Computer Vision,Yolov5,YOLO,object detection,AI,ML,tensorflow,keras,pytorch,artificial intelligence,machine learning,open cv,detection,tracking,data science
Tools: python,keras,tensorflow,pytorch,darknet
Department: Department of Computer Science
Project Team Members
| Name |
Email |
|
Qazi Arsalan Shah
|
qarsalan13@gmail.com |
Project Poster
