Responsibilities
- Design and refine models for detecting and tracking players and balls in sports video footage.
- Enhance solutions that align camera views with standardized field coordinate systems.
- Evaluate current computer vision workflows to set performance baselines and pinpoint areas for enhancement.
- Strengthen tracking accuracy under varying conditions such as stadiums, camera setups, and video quality.
- Plan and execute experiments spanning data collection, dataset development, augmentation, training, and evaluation.
- Investigate causes of model failures and implement fixes to boost precision and dependability.
- Modify existing systems to accommodate new sports, leagues, and camera configurations.
- Collaborate with data specialists on annotation processes and dataset validation.
- Coordinate with software and DevOps teams to transition models into live production systems.
- Optimize model inference speed, scalability, and operational monitoring.
- Define performance benchmarks and quality assurance procedures to maintain consistency over time.
- Lead projects from initial research and prototyping to deployment and iteration.
- Inform architectural choices that connect computer vision, machine learning, backend systems, and user workflows.
- Explain technical decisions and tradeoffs to engineering teams, clients, and leadership.
Work Arrangement
Remote (Worldwide)
Team
Remote-first environment with global collaboration
