DriveCast provides real-time object tracking and pose estimation from standard video streams using computer vision. The platform enhances these video feeds with Unreal Engine-powered computer graphics to create immersive augmented and virtual reality experiences.
Funding
Funding not disclosed
Founders
Product
Problem
Traditional autosport broadcasts struggle to maintain audience engagement due to outdated content formats that fail to provide comprehensive, real-time information about the race. Existing broadcast formats lack unique elements, making it difficult for content owners to stand out. New audiences find it challenging to understand the nuances of the sport without real-time guidance and context.
Solution
DriveCast offers a solution to enhance autosport broadcasts by using AI to track fast-moving objects in real time and overlay them with computer graphics. The platform uses standard video streams and integrates them with Unreal Engine to create augmented reality (AR) experiences. The system only requires six photos from different perspectives to train the AI to detect a specific car, and one qualifying session to test the concept. The AR layer provides graphic output tracked to the car, including racer, team, car details, interactions between cars, and checkpoint progress in rally racing.
Target Audience
DriveCast targets content owners, racing series, and online/TV broadcasters seeking to enhance their autosport broadcasts, as well as both beginner and advanced autosport enthusiasts.
Features
- AI-powered object tracking for fast-moving vehicles
- Real-time overlay of computer graphics onto video feeds
- Compatibility with any professional camera on track
- Remote operation requiring only 1-2 specialists
- AR visualizations of racer, team, car, and in-race interactions