It has helped us in getting better results with very less time compared to manual method of labeling videos. It also helps in automating repetitive task which can be performed by human eye but not possible without this tool! There are few features that I feel could have been improved upon like we cannot upload our own models or use any other custom trained network as it does support only pre-trained networks provided within its app/website. As well there's no option available to do semi supervised learning i e using some portion manually labeled images from your existing annotated set. This product solves majority part of problems faced while working with Videos - whether they are coming from different sources (recorded through webcams), recorded through mobile devices etc., so far so good! We were looking forward when someone comes up with such kind products where one doesn't need to pay hefty amount just to get quality work done from them. Some small things here & there would help me improve my skills.
ο»Ώ