Object Tracking Annotation

Tracking people on the streets

• 20,000 frames
• 4 weeks

Tracking each person in the frame by Bounding-Box markup using CVAT. The labeling is presented in the form of coordinates in XML format

APPLICATION AREAS

– Tracking the movement of people to analyze flows in commercial spaces in order to optimize the placement of goods

– Recognition of suspicious human behavior to alert security services and prevent crimes

– Crowd density assessment for monitoring at public events to ensure safety

– Tracking and automated counting of people to assess visitors to public spaces

– Tracking of workers at production to ensure labor safety and compliance with local regulations

Tracking a basketball

• 10,000 frames
• 2 weeks

Ball detection using Bounding-Box. Classification by overlap and position of the ball in CVAT. The labeling is presented in the form of coordinates and attributes in XML format

APPLICATION AREAS

– Ball tracking to monitor game statistics and analyze team performance

– Classification of ball position to determine the strengths and weaknesses of the team and individual players

– Ball tracking to predict game results and player statistics

– Ball tracking to help analyze players’ performance and help improve their skills

– Tracking to create statistical reports that can be used by coaches and team managers to make decisions

– Tracking and classification of ball position for use in training for children and beginning players to help them develop and improve their skills

Tracking cars on video

• 20,000 frames
• 4 weeks

Tracking each car in the frame using Bounding-Box. Classification of cars depending on the type of body. The labeling is presented in the form of coordinates and attributes in XML format

APPLICATION AREAS

– Identification and tracking of vehicle speed to ensure safety on the highways

– Traffic tracking and counting to monitor road congestion in real time

– Classification of vehicles by body type for traffic analysis of urban infrastructure

– Recognition of abnormal vehicle behavior to alert security services in real time