Staff Machine Learning Engineer
About The Position
XPENG is a leading smart technology company at the forefront of innovation, integrating advanced AI and autonomous driving technologies into its vehicles, including electric vehicles (EVs), electric vertical take-off and landing (eVTOL) aircraft, and robotics. With a strong focus on intelligent mobility, XPENG is dedicated to reshaping the future of transportation through cutting-edge R&D in AI, machine learning, and smart connectivity. About the Role We are looking for a strong Machine Learning Engineer / Computer Vision Engineer to work on Traffic Sign Recognition (TSR) 2D detection for production autonomous driving systems. In this role, you will be responsible for the full lifecycle of TSR model development, including scenario analysis, data preparation, model training, evaluation, optimization, quantization, and deployment. You will work closely with perception, data, infrastructure, and deployment teams to improve traffic sign detection performance across diverse real-world driving scenarios. This role is ideal for candidates who enjoy solving practical computer vision problems, building reliable model iteration pipelines, and bringing perception models from offline training to onboard production systems.
Requirements
- Master’s, or PhD degree in Computer Science, Electrical Engineering, Robotics, Computer Vision, Machine Learning, or a related field.
- 3-5 years of strong hands-on experience with computer vision models, especially object detection.
- Experience with detection architectures such as YOLO, Faster R-CNN, DETR/Deformable DETR, RT-DETR, RTMDet, or similar models.
- Proficiency in Python and deep learning frameworks such as PyTorch or TensorFlow.
- Solid understanding of object detection training workflows, including dataset preparation, augmentation, loss functions, evaluation metrics, and model debugging.
- Experience with common detection metrics such as mAP, precision/recall, false positive/false negative analysis, and class-level performance breakdown.
- Strong data analysis and problem-solving skills.
- Ability to work cross-functionally with model, data, infrastructure, and deployment teams.
Nice To Haves
- Experience in autonomous driving, ADAS, robotics, or safety-critical perception systems.
- Experience with traffic sign recognition, traffic light recognition, road object detection, or small-object detection.
- Familiarity with long-tail scenario mining, hard negative mining, class imbalance handling, and dataset curation.
- Experience with ONNX, TensorRT, model quantization, C++ inference pipelines, CUDA, or edge deployment.
- Experience debugging training-to-deployment consistency issues, including preprocessing mismatch, postprocessing mismatch, quantization accuracy drop, or runtime performance bottlenecks.
- Familiarity with large-scale data pipelines, scenario tagging, or automated data mining workflows.
- Strong engineering discipline in experiment tracking, reproducibility, regression testing, and model version management.
Responsibilities
- Develop and improve 2D traffic sign detection models for autonomous driving perception systems.
- Analyze TSR-related scenarios and failure cases, including missed detections, false positives, occlusions, small objects, rare signs, region-specific signs, and adverse weather or lighting conditions.
- Prepare, clean, curate, and analyze training and evaluation datasets for TSR model iteration.
- Design and execute model training experiments, including data sampling, augmentation, loss tuning, class imbalance handling, and hard-case mining.
- Build and maintain evaluation pipelines for TSR models, including offline metrics, scenario-based evaluation, regression testing, and error analysis.
- Collaborate with data teams to define mining strategies for long-tail TSR scenarios and improve dataset coverage.
- Optimize models for production deployment, including ONNX / TensorRT / quantization / inference acceleration.
- Work with deployment and platform teams to validate model performance on onboard or edge compute platforms.
- Track model performance across versions and support continuous improvement through data-model-evaluation feedback loops.
- Debug issues across the full stack, including data quality, labeling, model behavior, evaluation mismatch, and deployment consistency.
Benefits
- Competitive compensation package.
- Snacks, lunches, dinners, and fun activities.
- Infrastructures and computational resources to support your work.
- Opportunity to work on cutting edge technologies with the top talents in the field.
- Opportunity to make a significant impact on the transportation revolution by the means of advancing autonomous driving.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
Career Level
Senior
