[latexpage]
Teknikimobil.com – Mighty AI announced a series of new product offerings available in its data management platform for autonomous vehicle perception data. These new features are intended to help customers better manage ground truth data for developing machine learning models for autonomous vehicles. They include blurred faces and license plates for privacy, gaze tracking workflows to improve driver safety, and portals for managing autonomous vehicle data. This enhancement complements Mighty AI’s cloud-based annotation platform to generate ground truth datasets to improve vehicle perception models.
Also read:
5 Ways Artificial Intelligence is [Driving] Progress in the Automotive Industry
Elon Musk’s Neuralink Biological Research Laboratory – Maybe it will be built
Photo by mariordo59
The first offering involves privacy protection ahead of GDPR (General Data Protection Regulation), a new European Union regulation designed to strengthen data protection for individuals. The GDPR, entered into force May 25, 2018.
To protect the privacy of individuals appearing in raw image data companies collect throughout the development cycle of autonomous vehicles. Mighty AI launched a privacy solution, which includes facial automation and blurred license plates.
The auto-blur functionality aims to anonymize people and their vehicles, allowing the company to hide personally identifiable information before using the data in its development cycle. In addition to this new technology, the Mighty AI annotation community must also sign a non-disclosure agreement before accessing images to further enhance data privacy.
Mighty AI Founder and CEO Daryn Nakhuda said that whenever you are dealing with data that contains personal information, privacy is a concern. He added that autonomous vehicles collect huge amounts of real-world data that must be labeled to train perception systems. An automatic blur solution that aims to anonymize individuals in such data is a major step towards protecting the privacy of every vehicle owner.
Combination of Open Source Internal Data Sets
Offers using a combination of internal and open source data sets. Mighty AI trains a proprietary, purpose-built deep neural network to detect faces and license plates in AV scenarios. Today, Mighty AI uses models to automatically apply an irreversible blurring process to detected faces and license plates.
Another offering from Mighty AI, eye gaze, was developed with the aim of improving driver safety. Computer vision systems can determine a driver’s ability to assume control of an autonomous vehicle. The method is to detect and monitor a person’s face, expression and gaze. In this way, the system can detect whether the driver is at risk of falling asleep.
Mighty AI introduces the tools necessary for its customers to develop these important driver safety systems. An idea that is particularly relevant for designing Level 2 and Level 3 autonomous vehicles. The new tooling generates keypoint annotations to label dozens of different points on the human face, which are then used by customers to develop computer vision systems to track these points and determine concerns driver and readiness to take control of an autonomous vehicle.
Also read:
Arilou Unveils New Automotive Ethernet Technology
Tesla’s AI-Based Autopilot 2.0 Tested After Model S Owners Conduct Experiments
Finally, Mighty AI announced tools to simplify autonomous vehicle data management. Called Mighty Studio, it is a portal for customers to manage their ground truth datasets. Mighty Studio performs quality assurance checks of ground truth datasets against objective measures. It offers features such as filtering by object class and date range, and the ability to view annotations.[]