Full Job Description
Develop, train, and evaluate computer-vision models for autonomous-driving perception tasks, applying both classical computer-vision methods and deep-learning approaches. Design and conduct controlled machine-learning experiments, develop quantitative evaluation metrics, and analyze results to improve model and dataset quality. Develop and maintain pipelines for scene mining, annotation-task generation, dataset construction, data processing, model training, and validation. Develop software tools for analyzing model behavior and visualizing annotations, datasets, predictions, and evaluation results. Develop, integrate, and maintain computer-vision components within offline, simulation, and online perception systems, including multicamera and panoramic camera configurations, and resolve image-projection, geometric-transformation, and data-synchronization issues arising across those systems. Review scientific literature in computer vision and deep learning and prototype and evaluate relevant models and algorithms for autonomous-driving perception problems. Collaborate with related engineering teams to investigate data, annotation, infrastructure, and cross-component integration issues.
Requirements:
Master's degree in Computer Science, Computer Engineering, Electrical Engineering, Applied Mathematics, Mathematics, Informatics, or a closely related field. A foreign equivalent degree is accepted. Position requires 36 months of experience as a research engineer, machine-learning engineer, or position in a related field of computer vision or machine-learning model development.
The position further requires the following:
At least one year of that experience must have involved developing camera-based perception systems for autonomous vehicles or advanced driver-assistance systems (ADAS).
Two years of the required professional experience must have involved developing and evaluating deep-learning-based computer-vision models.
Two years of the required professional experience must have involved designing or conducting machine-learning experiments and evaluating model performance.
Experience with classical computer vision, image processing, object detection, segmentation, and tracking.
Experience developing data-processing, annotation, dataset-generation, training, or validation pipelines for machine learning.
Professional software-development experience in Python and C++.
Experience with PyTorch or TensorFlow.
Experience with distributed or large-scale data processing using PySpark or a comparable framework.
Experience with workflow orchestration for data-processing or model-training pipelines using Argo Workflows or a comparable system.
Experience developing in containerized environments using Docker or a comparable technology.
Experience using SQL to query and prepare data for analysis, model training, or evaluation.
Experience with distributed version control using Git or a comparable system.
Candidates are required to be authorized to work in the U.S. The employer is not offering relocation, sponsorship, and remote work options are not available.