Tagging the Make and Models of Cars in Videos
Published date: 19.03.2021
Read time: 1 min
Client Profile
Industry: Technology services
Location: Slovenia
Size: 11-50 employees
Company Bio
The client offers out-of-box intelligent security solutions on the EDGE (IP video camera processor), that support traffic, access control, and parking management projects. They build their products using vehicle plate, make, model, color recognition; intellectual data processing; “camera-to-camera” communication technologies.
Business Problem
The client was working on a product that would use computer vision to distinguish between all of the various makes and models of cars. In order to further train the machine learning algorithms, they needed to annotate a dataset of 50,000 short videos that were taken in various parking lots. Since this is a very time-consuming task, they turned to Mindy Support for assistance.
Solutions Provided by Mindy Support
Mindy Support assembled a team of 18 full-time data annotators to perform the annotation work. The amount of classifications for each vehicle was 750 classes in the most complex cases. Nevertheless, our team got the job done in 15 days, which was the timeframe discussed with the client with a quality rate of >95%.
Results Delivered to the Client
GET A QUOTE FOR YOUR PROJECT
We have a minimum threshold for starting any new project, which is 735 productive man-hours a month (equivalent to 5 graphic annotators working on the task monthly).
MORE SUCCESS STORIES
FOLLOW US ON OUR SOCIAL NETWORKS
