LeddarTech Holdings and Immervision have announced a collaboration that aims to streamline the perception model training process to make it faster, more cost-effective and less data-intensive. LeddarTech is an automotive software company that provides patented disruptive AI-based low-level sensor fusion and perception software technology, LeddarVision™, for advanced driver assistance systems (ADAS) and autonomous driving (AD) and parking applications. Immervision is a leading developer of advanced vision systems combining optics, image processing and sensor fusion technology.
This joint effort combines LeddarTech’s expertise in AI-based low-level sensor fusion and perception software technology and Immervision’s leadership in advanced vision systems, creating a powerful synergy intended to improve ADAS and AD perception training.
The goal of perception technology in ADAS is to provide accurate and timely information to the vehicle’s control system, enabling it to make informed decisions and assist the driver in various driving tasks, such as collision avoidance, lane keeping and adaptive cruise control. However, the development and validation process requires costly data collection and labelling.
LeddarTech and Immervision’s collaboration aims to improve and significantly optimize this process by allowing large-scale data re-usability. The breakthrough achieved by their research teams allows the re-use of data across various ADAS and AD platforms, with zero simulation-to-reality gap. This means that existing data assets can now be leveraged for new platforms, significantly shrinking the development time and cost for the perception models, thus enabling a faster and more cost-effective development cycle for the complete system. This optimization is expected to result in a more cost-effective and safer future for ADAS and AD.