Lucid's ADAS/Autonomous Driving division needs to enhance the efficiency of deep learning models for real-time inference.
Requirements
- Strong experience in CUDA kernel development and TensorRT plugin optimization for deep learning inference
- Proficiency in C/C++ programming, particularly for embedded systems and real-time applications
- Solid understanding of deep learning model architectures, with hands-on experience optimizing models for deployment
- Familiarity with automotive safety standards (e.g., ASPICE, ISO 26262)
- Experience working with automotive sensors (e.g., Camera, Radar, Lidar) in ADAS/AD applications
- Knowledge of Neural Architecture Search (NAS) techniques for optimizing deep learning model architectures
Responsibilities
- Analyze ADAS/AD hardware to identify deep learning model optimization opportunities
- Lead the technical roadmap for deep learning inference optimization
- Develop and integrate custom model optimizations using internal datasets and benchmarks
- Debug and enhance deep learning deployment pipelines
- Conduct unit tests and validation to ensure the reliability, accuracy, and efficiency of optimized models
- Collaborate with perception, software, and hardware teams to ensure optimized models meet real-time performance requirements
Other
- Bachelor's degree in Computer Engineering, Electrical Engineering, Automotive Engineering, Mechanical Engineering, or a related field
- Minimum 3 years of professional experience or a Ph.D. for senior positions
- Advanced degrees preferred
- Familiarity with agile development methodologies and collaborative software development
