Bosch Group
智能驾驶数据科学家_CR
Salary
Job description
- 研发面向自动驾驶、以数据为中心的算法，重点包括多模态数据（camera, lidar, radar）上的 auto annotation、auto tagging、data mining 和 auto quality check。
- 利用 Foundation Models、VLM、few-shot learning 和 zero-shot learning 来开发和优化数据自动化流水线，以提高效率和可扩展性。
- 实施 active learning 策略，进行智能数据选择和标注优先级排序。
- 与感知、预测和规划团队协作，理解数据需求并提供可扩展的数据解决方案。
- 与全球博世团队合作，进行技术转移、趋势追踪和方案评估。
- Research and development of data-centric algorithms for autonomous driving, focusing on auto annotation, auto tagging, data mining, and auto quality check across multi-modal data (camera, lidar, radar).
- Develop and optimize data automation pipelines leveraging Foundation Models, VLM, few-shot learning, and zero-shot learning to improve efficiency and scalability.
- Implement active learning strategies for intelligent data selection and annotation prioritization.
- Collaborate with perception, prediction, and planning teams to understand data requirements and deliver scalable data solutions.
- Work with global Bosch units on technology transfer, trend scouting, and concept evaluation.
- 计算机科学、电气工程、数据科学或相关专业的硕士或博士学位。
- 拥有1-3年在自动驾驶或AI应用领域担任以数据为中心角色的实践经验。
- 精通 Foundation Models (VLM, e.g., CLIP, Grounded-SAM, Grounded-DINO)、few-shot/zero-shot learning 和 active learning。
- 具备 auto annotation、data mining 和 auto quality check 流水线的实践经验。
- 熟练掌握 Python 及 PyTorch 或 TensorFlow 等深度学习框架。
- 熟悉多模态传感器数据（cameras, lidar, radar）。
- 在顶级会议（如 CVPR、ICCV、ECCV）以第一作者身份发表论文者优先。
- 加分项：具有使用 large feed-forward models 完成 3D reconstruction、depth estimation 或预测 camera intrinsic/extrinsic matrix 等任务的经验。
- 英语流利，具备强大的沟通和团队合作能力。
- Master’s or PhD in Computer Science, Electrical Engineering, Data Science, or a related field.
- 1-3 years of practical experience in data-centric roles within autonomous driving or AI applications.
- Strong knowledge of Foundation Models (VLM, e.g., CLIP, Grounded-SAM, Grounded-DINO), few-shot/zero-shot learning, and active learning.
- Hands-on experience with auto annotation, data mining, and auto quality check pipelines.
- Proficiency in Python and deep learning frameworks such as PyTorch or TensorFlow.
- Familiarity with multi-modal sensor data (cameras, lidar, radar).
- First-author publication at top conferences (e.g., CVPR, ICCV, ECCV) is a strong plus.
- Bonus: Experience with tasks such as 3D reconstruction, depth estimation, or predicting camera intrinsic/extrinsic matrix using large feed-forward models.
- Fluent in English, with strong communication and teamwork skills.
