Data annotation and labeling for the development of driver
assistance systems
Data annotation (also known as labeling) is an essential process for machine learning and algorithm development. By adding information and categories to raw data from cameras, radar or lidar sensors the data can be used to train AI models or verify algorithms
Our coordinated onshore and nearshore capabilities enable us to deliver large, high-quality projects at attractive prices. We use project-specific tools which provide the opportunity for immediate feedback and short development iterations for key features
Our quality promise for your project:
1. Requirements Management:
We collaborate with your engineers to
release a well aligned labeling reference guide to ensure that all
requirements are implemented later accurately and efficiently.
2. Alignment of tools and custom interfaces:
We ensure the seamless
integration of project-specific tools and interfaces, tailored to your
needs and the requirements of the project.
3. Customised data exchange:
We manage the transfer and conversion
of your data to ensure a smooth and efficient exchange.
4. Start of annotation:
After careful preparation, we start the annotation
process with focus on high data quality from day one.
5. Regular quality assurance reviews:
We conduct regular reviews with
your team to ensure that all project requirements are met and that
quality is maintained throughout the process.
6. Project specific reporting:
We provide you with customized reports
that give you a regular overview of the progress of the project and the
quality of the work.
7. Completion & handover:
Upon successful project completion, we
ensure that all deliverables are fully available and that the delivery
process runs smoothly.
8. Retrospectives and continuous process improvement:
We conduct
retrospective validations to learn from each project and constantly
optimize the entire process to make future projects even more
efficient.
Using AI for railway safety
Since February 2023, we have been working with FusionSystems GmbH on a
project of the German Centre for Rail Traffic Research (DZSF) on the use of
artificial intelligence in automated driving in railway transport. The aim of
the project is to develop general guidelines for the preparation of data for
training and validation of AI algorithms for object recognition and
classification, as well as to prepare AI data for a subset of the relevant input
parameters and to publish them at the end of the project. The results of the
project should support the research and development of applications for
object recognition and classification in automated driving in railway
operations.
More information can be found here:
