What the client needs
To gain a competitive edge in the autonomous vehicle race, the demand for high-quality data annotation services for AI models is rapidly increasing. So, it is one of the main reasons why our clients are seeking a data annotation vendor capable of delivering their data annotation project results with an accuracy rate of 98% and above.
How we did it
With an 8-year track record in data annotation for the ADAS industry, we swiftly assembled a team of 38 professionals, including project managers, reviewers, and annotators. The recruitment process was completed within 1-2 weeks, followed by an intensive training program to meet strict project deadlines. Thanks to the team’s seamless collaboration, we successfully annotated 26,000 images within six months.
In this project, we used panoptic segmentation to annotate over 40 data categories, including vehicles, pedestrians, obstacles, traffic signs, traffic lights, roads, sky, and more. After a training period, our team members continuously improved their quality and completed the project with impressive results. We not only finished the project ahead of schedule but also impressed the client with an error rate of less than 1%. The best news is that the client has invited us to participate in the next phase of the project’s expansion.
What the results have
- 26,000 images annotated in 6 months
- Achieved an accuracy rate of 99.4%
- Used panoptic segmentation technique to annotate over 40 data categories
