We are seeking a highly skilled and motivated AI/ML Researcher to design, develop, and deploy cutting-edge computer vision systems specifically for driver modeling applications. This role is pivotal in enhancing the safety, comfort, and personalization of our next-generation automotive interior monitoring products. The ideal candidate will combine strong research skills with practical engineering experience to translate innovation into production-ready, robust, and reliable driver modeling solutions for the automotive industry.
Key Responsibilities:
- Research & Development for Driver Modeling: Conduct pioneering research to develop novel algorithms and models specifically tailored for critical driver modeling tasks, including:
- Domain-Aware Representation Learning: Investigate and develop effective methods for extracting robust, semantically meaningful, and interpretable features from multi-modal sensor data. This includes leveraging geometric, behavioral cues, alongside modern representation learning techniques, to capture essential aspects of driver state actions and behavior.
- Driver State and Behavior Analysis: Advanced detection of driver attributes like Distraction, Drowsiness (including micro-sleeps), Sleep onset, and emotional states. This involves developing sophisticated models to understand and predict complex human behavior, leveraging statistical methods, information theory, and probabilistic approaches.
- Driver Intent and Activity Recognition: Interpreting driver intent, head pose, gaze tracking, eye- gaze estimation, and recognizing specific activities (e.g., phone usage, eating, interacting with vehicle controls).
- 3D Interior Scene and Driver Interaction Understanding: Interpreting the driver's environment and their interactions within the vehicle cabin for a holistic understanding.
- Algorithm Implementation & Optimization for Automotive: Design, train, and optimize deep learning models (e.g., CNNs, Transformers, Recurrent Neural Networks) and advanced traditional computer vision algorithms for core perception tasks central to driver modeling, utilizing multi-modal sensor data commonly found in automotive environments (e.g., RGB, IR, Depth cameras). Focus on computational efficiency and real-time performance suitable for embedded systems.
- Automotive-Grade Data & Pipelines: Devise comprehensive data collection, annotation, and augmentation strategies specifically for diverse driver populations and real-world driving scenarios. Manage and curate large-scale, high-quality automotive datasets for model training, validation, and benchmarking.
- Stay abreast of the latest advancements in AI/ML and computer vision research relevant to driver modeling, contributing to the company's IP through patents or publications.
Required Qualifications & Skills:
- Education: Master's or PhD in AI/ML, Computer Science, Electrical Engineering, or a related technical field, or equivalent practical experience with a strong focus on computer vision for automotive applications.
- Technical Proficiency: Strong programming skills in Python and C++, with experience in optimizing code for performance.
- ML/CV Expertise: Deep understanding of computer vision fundamentals, image processing, and proven experience with both traditional computer vision techniques (e.g., feature descriptors, geometric vision, optical flow) and machine learning/deep learning frameworks (such as PyTorch and TensorFlow). Ability to intelligently combine these approaches for robust solutions.
- Hands-on Driver Modeling Experience: Demonstrable hands-on experience building, training, and deploying perception systems specifically for driver modeling, human-centric AI, or related automotive interior sensing applications. This includes academic projects, industry experience, or contributions to open-source initiatives focused on driver understanding.
- Core Concepts: Solid grasp of mathematical, statistical, and information theory concepts, including linear algebra, calculus, probability theory, stochastic processes, and concepts like entropy. This theoretical foundation is essential for advanced algorithm development, particularly in modeling complex human behavior and uncertainty.
- Problem-Solving: Excellent analytical and problem-solving skills, with the ability to tackle complex, novel challenges inherent in developing safety-critical automotive systems.
Preferred Qualifications:
- Publications: Publications at top-tier AI/CV conferences (e.g., CVPR, ICCV, ECCV, NeurIPS) related to human pose estimation, facial analysis, gaze tracking, activity recognition, feature engineering for behavioral cues, or driver behavior analysis.
- Multi-modal Sensor Fusion: Experience with multi-modal sensor fusion techniques (e.g., camera, RGB/IR, radar) for robust perception in automotive environments.
- Embedded Systems/Optimization: Experience with model compression, quantization, and deployment on embedded automotive hardware or resource-constrained devices.
- Real-time Systems: Experience designing and optimizing algorithms for real-time performance in production environments.