Intern_ADAS VLA Intern/智驾大模型实习
招聘时间
即日起 至 2026-12-31
招聘单位
梅赛德斯-奔驰
所在地区
北京
工作地点
北京市/北京市/朝阳区
学历要求
本科
实习时长
每周5天，每天6小时
职责描述
[6 months, 4-5 days/week]
Main Tasks:
Perception pipeline development (30%)
- Support to develop ADAS perception and fusion pipeline that utilize data from various sensors, inlcuding camera, lidar, radar, and ultrasonic sensors
- Support to develop perception auto-labeling pipelines and continously opitmizing them with better model architecture and evaluation metrics
Advanced Algorithm Research and Application (40%)
- Research, benchmark, and deploy advance Vision-Language-Action (VLA) and Vision-Language Model (VLM) algorithms with MB datasets
- Continously explore and integrate advanced models to improve system understanding and overall performance in diverse and challenging driving and parking scenarios
Cross-Functional Integration and Collaboration (30%)
- Work closely with AD research and engineering teams to integrate SOTA methods to internal ADAS architecture
- Collaborate with data, validation and product team stakeholders to align development with data needs, ensure systematical functional evaluation and guarantee successful final delivery
Key Qualifications:
1. In-depth knowledge of computer vision, machine learning, deep learning, sensor fusion and robotics
2. Deep understanding of state-of-the-art VLM/LLM/VLA algorithms
3. Familiar with machine learning and deep learning workflow, including data preparation, model design, training and evaluation metrics design
4. Familiar with Tensorflow and Pytorch framework, familiar with ROS
5. Strong coding skills in Python/C++
6. Good communication skills and teamwork ability
7. Fluent English language skills preferred
Language:
Chinese: Conversational
English: Conversational
Education:
Degree: Bachelor or above
Major: Computer Science, Electronic Engineering
Main Tasks:
Perception pipeline development (30%)
- Support to develop ADAS perception and fusion pipeline that utilize data from various sensors, inlcuding camera, lidar, radar, and ultrasonic sensors
- Support to develop perception auto-labeling pipelines and continously opitmizing them with better model architecture and evaluation metrics
Advanced Algorithm Research and Application (40%)
- Research, benchmark, and deploy advance Vision-Language-Action (VLA) and Vision-Language Model (VLM) algorithms with MB datasets
- Continously explore and integrate advanced models to improve system understanding and overall performance in diverse and challenging driving and parking scenarios
Cross-Functional Integration and Collaboration (30%)
- Work closely with AD research and engineering teams to integrate SOTA methods to internal ADAS architecture
- Collaborate with data, validation and product team stakeholders to align development with data needs, ensure systematical functional evaluation and guarantee successful final delivery
Key Qualifications:
1. In-depth knowledge of computer vision, machine learning, deep learning, sensor fusion and robotics
2. Deep understanding of state-of-the-art VLM/LLM/VLA algorithms
3. Familiar with machine learning and deep learning workflow, including data preparation, model design, training and evaluation metrics design
4. Familiar with Tensorflow and Pytorch framework, familiar with ROS
5. Strong coding skills in Python/C++
6. Good communication skills and teamwork ability
7. Fluent English language skills preferred
Language:
Chinese: Conversational
English: Conversational
Education:
Degree: Bachelor or above
Major: Computer Science, Electronic Engineering
相关实习岗位
