VAYAVISION, a LeddarTech company,  brings autonomous vehicle perception to new heights. By applying our AI and computer vision algorithms to raw data fused from radar, lidar and camera, we build the most accurate 3D environmental model.

Most Accurate Environmental Model

Deep learning and computer vision algorithms are applied to the upsampled high-resolution colored 3D model (RGBd), resulting in remarkably accurate and reliable detection and classification.

All information pertaining to the size, shape and velocity of every surrounding object, including small obstacles, can be viewed and analyzed.

Deep Tech Raw Data Fusion with Upsampling

Lidar, Radar and camera are perfectly calibrated and synchronized to receive exact pixel-level matching. Sparse low-resolution data samples from the lidar and radar images are upsampled to HD quality, enabling to assign depth and speed information to every pixel resulting in HD RGBd model.


With Better

VAYAVISION excels at detecting even unknown objects that are absent from training data set.

Our novel approach comes with Inherent functional safety, detecting objects and dangers even during a sensor malfunction.

Best AV Perception

At Low System Cost

Cost-effective fusion-based solution with upsampling has low sensor requirements and low computational requirements. Low-cost depth sensors with low resolution suffice to deliver most reliable environmental perception