ITRI has integrated computer vision, deep learning, and sensor fusion technology to build a highly cost-effective AI advanced driver assistance systems (AI-ADAS) solution. When embedded within a platform, this solution can realize, integrate and optimize seven types of detection function (traffic lights, lane markings, drivable area, four-wheel vehicles, twowheel vehicles, pedestrians, signal status and semantics). When combined with deep learning, AI-ADAS can achieve greater effectiveness. The Lane Departure Warning System (LDWS) and Forward Collision Warning System (FCWS) have been introduced for commercial applications.
ITRI has also collaborated with vendors to develop AI-DVR products for emerging smart traffic applications, as well as AI-ADAS technology to differentiate products. Furthermore, ITRI has linked different industries in Taiwan, such as the semiconductor, IC design, module, subsystem, and car manufacturing industries to construct a development and verification ecosystem for vehicle-mounted deep-learning edge computing devices. The goal is to promote the development of new and increased-value products among Taiwan’s information and communication technology (ICT) vendors.
Applications & Benefits
- Realizing, integrating, and optimizing seven types of detection function - traffic lights, lane markings, drivable area, four-wheel vehicles, twowheel vehicles, pedestrians, signal status, and semantics.
- Achieving greater effectiveness when combining with deep learning.