Skip to content

StreamrCVAT/streamrcvat

Repository files navigation

Semi-Automated Video Annotation tool using Deep Learning

Overview

The objective of this product is to allow automatic annotation objects in a video by using AI algorithms to reduce the overall time and cost consumed in the annotation process.

Dataset

Dataset URL : https://youtu.be/GkyJmcS2EcA
Dataset description : A grey car moving across the camera
Duration of the video : 10 seconds
Number of frames split : 500
Annotation Tool used : LabelImg
Annotation Tool URL : https://github.com/tzutalin/labelImg
Annotated Image format : YOLO (Coordinates in .txt file)
Average time to annotate 1
Frame : 12.84 seconds/frame

Other datasets

https://www.jpjodoin.com/urbantracker/dataset.html

Core contributors