Perception Model Engineer

Key Responsibilities:
Handle autonomous driving data collection, cleaning, annotation, quality checking, and data loop construction.
Participate in annotation specification design and improve annotation tools and workflows.
Conduct algorithm selection, research, reproduction, and benchmarking for intelligent driving.
Train, evaluate, fine-tune, and optimize deep learning models.
Iterate and optimize intelligent driving algorithms for accuracy, robustness, latency, and resource efficiency.
Participate in algorithm research and design, track cutting-edge technologies, and produce technical documentation.
Collaborate with perception, planning/control, engineering, product, and data teams to deploy algorithms.
Requirements:
Master’s degree or above in Mathematics, Physics, Computer Science, Electronic Engineering, Robotics, Artificial Intelligence, or related fields.
Fresh graduates: no prior work experience required, but solid fundamentals in mathematics, Python, and deep learning are expected.
Experienced candidates: solid Python programming and deep learning fundamentals; familiarity with at least one mainstream framework such as PyTorch.
Strong logical thinking, communication skills, and a high sense of responsibility.
Good English working proficiency for cross-cultural and cross-functional collaboration.
Knowledge of intelligent driving, autonomous driving, computer vision, point cloud, sensor fusion, SLAM, planning/control, or related areas is a plus.
Nice to Have:
Experience in autonomous driving, ADAS, or robotics projects.
Familiarity with KITTI, nuScenes, Waymo, CARLA, ROSbag, or similar datasets/tools.
Experience with C++, Linux, ROS, Git, Docker, or CUDA.
Publications, competitions, open-source contributions, or patents.

Work arrangement
Yes