某知名自动驾驶公司
感知算法工程师
信息技术
科技
国际
3-5年
本科
¥50 - 60K/月
职位描述
Responsibilities
1. Perception Algorithm Development
Design and develop core perception algorithms for autonomous driving, including object detection, tracking, classification, semantic segmentation, 3D perception, and scene understanding.
Work with data from camera, LiDAR, radar, and other sensors to build robust perception modules.
2. Multi‑Sensor Fusion
Develop and optimize multi-sensor fusion algorithms to improve perception accuracy, robustness, and environmental adaptability.
Implement fusion methods such as Kalman filters, Bayesian inference, and deep-learning-based fusion.
3. Algorithm Optimization & Deployment
Optimize algorithms for real-time performance, memory efficiency, and stability to meet automotive-grade requirements.
Deploy algorithms on embedded platforms or high-performance computing systems; perform profiling and acceleration using CUDA, TensorRT, or similar tools.
4. Data Analysis & Iteration
Participate in data collection, annotation, and analysis.
Build data-driven pipelines for continuous algorithm improvement, especially for challenging scenarios such as night, rain, fog, occlusion, and long-tail cases.
5. Research & Innovation
Track cutting-edge research in computer vision, deep learning, and autonomous driving.
Evaluate and integrate promising technologies into production systems.
6. Cross‑Functional Collaboration
Collaborate closely with planning, control, localization, HD map, and system engineering teams.
Ensure seamless integration of perception modules into the full autonomous driving stack.
职位要求
Qualifications
1. Educational Background
Master’s degree or above in Computer Science, Electrical Engineering, Robotics, Machine Learning, Artificial Intelligence, or related fields (exceptional bachelor’s candidates considered).
2. Technical Skills
Strong foundation in mathematics and algorithms, including linear algebra, probability, optimization, and numerical methods.
Solid understanding of computer vision and deep learning algorithms (e.g., YOLO, Faster R‑CNN, CenterPoint, BEV models, UNet, DeepLab).
Proficiency in C++ and Python with strong engineering and debugging skills.
Experience with deep learning frameworks such as PyTorch, TensorFlow, or JAX.
Familiarity with GPU programming, CUDA optimization, or inference engines (TensorRT, ONNX Runtime) is a plus.
3. Sensor & Domain Knowledge
Understanding of camera, LiDAR, radar, and other sensor principles and data characteristics.
Experience in autonomous driving, ADAS, robotics, or related perception systems is preferred.
4. Research & Project Experience (Bonus)
Publications in top-tier conferences/journals (CVPR, ICCV, ECCV, NeurIPS, ICRA, etc.) are a strong plus.
Experience with open-source perception frameworks (e.g., OpenPCDet, MMDetection3D) or participation in competitions (Waymo, nuScenes, KITTI) is advantageous.
