AI/ML ADAS Perception Engineer
Salary not listed • Auburn Hills, Michigan, US • Full-time
Posted 1mo ago
No salary is listed. Many US states require employers to post a pay range, so ask about pay early.
Job Location
Auburn Hills, Michigan, US
Tech Stack
Remote Work Policy
On-site
Employment Type
Full-time
Categories
Machine Learning Engineer
About the job
We are seeking a Computer Vision Algorithm Engineer focused on ADAS perception to transform camera video feeds into a clear understanding of the vehicle's surroundings. This role spans concept through serial production, involving algorithm design, simulation, data analysis, and the development and optimization of classical and deep learning models for real-time execution on production ECUs. Key tasks include object detection, segmentation, and tracking, all while meeting stringent accuracy, latency, memory, and power targets. A strong foundation in image processing, machine learning, and mathematics is essential, with familiarity in vehicle dynamics being a plus.
Responsibilities
- Develop, implement, and optimize conventional image processing algorithms for automotive embedded serial production.
- Design, develop, tune, and optimize deep learning models for ADAS computer vision features, focusing on pruning, quantization, and computational performance.
- Plan and execute experiments to assess deep learning model effectiveness, compare architectures, and validate results through rigorous testing.
- Analyze large datasets to extract insights, refine models, and improve overall performance and robustness.
- Incorporate advances in deep learning and research findings into projects.
- Collaborate with multidisciplinary teams to integrate models into existing systems.
- Document development processes, maintain experiment logs, and present findings to stakeholders.
- Analyze defects and test results, perform root-cause analysis, and implement algorithm improvements to achieve KPIs.
- Independently deliver ADAS algorithm design, implementation, and testing.
Requirements
- Master's degree in computer engineering, Software Engineering, Electrical Engineering, Computer Science, or equivalent.
- Minimum of 3 years of experience in computer vision and image-processing algorithm development using traditional methods and deep learning.
- Proven expertise in developing and implementing DNN models.
- Excellent programming skills with C or C++.
- Familiarity with Python and proficiency in deep learning frameworks (TensorFlow, PyTorch, Keras) is advantageous.
- Strong grasp of machine learning concepts and neural network architectures (CNNs, RNNs, transformers).
- Experience in image segmentation, object detection, and image data preparation/enhancement.
- Experience in optimizing models for performance, including quantization and distributed training.
- Strong problem-solving abilities, capable of working independently and collaboratively.
- Effective communication of complex concepts to technical and non-technical audiences.
- Experience with automotive product development practices and structured engineering methodologies.
- Strong troubleshooting and debugging skills using structured problem-solving methods.
- Experience with disciplined software development processes (ASPICE or CMMI), configuration management, and project monitoring/control techniques.
- Strong foundations in mathematics and signal/image/video processing.
- Computer vision fundamentals (object detection, tracking, feature extraction) with C/C++.
- Experience developing AI and Machine Learning Algorithms for embedded devices.