synmove

中文
Next.js logo
  • We are a technology integrator in the field of autonomous driving, bringing together Taiwan's top talent in science and engineering. With a passion for leveraging advancements in modern machine learning, we aspire to develop world-class autonomous driving software.
    • Real-time Image Semantic Segmentation
    ICNet[2017]
    Real-time LIDAR Semantic Segmentation
    Train SOTA 3D point cloud segmentation models on Real-World road scenes such as Kitti 360
    To get a high accuracy, fast and robust semantic segmentation model
    Segmentation result on a busy intersection
    Full Self Driving
    Multi-Cam CV Based FSD with HD Map -- NuPlan Offline Perception
    2DPASS[2022]
    Moreover, we can fuse camera info to assist LIDAR point cloud segmentation
    Source
    To get a accurate LIDAR x Cam FSD system
    Image Segmentation for HD Map