Computer Vision Engineer (US)

Key Skills

Python Programming, C++, Keras software library, PyTorch, TensorFlow, Scikit-learn, Computer Vision (CV)

Job Description

Second Spectrum is building the next way of seeing sports. We deliver unique and innovative solutions through unparalleled capabilities in machine understanding of sports.

• We are the official tracking and analytics provider of the English Premier League, NBA, and MLS and have numerous clients across multiple sports at the team, league, and media level.
• We developed a real-time video augmentation pipeline that powers the innovative content featured in products like the award-winning Clippers CourtVision, and has also been used by ESPN, Turner Sports, BT Sport, FOX, and International NBA League Pass.
• We produce the most accurate tracking data with the fastest high-quality output by leveraging the most automated and extensible solution in the market.

We are constantly innovating. We believe that technology will revolutionize the way that sports are played, coached, and experienced, and that the work to make this happen has just begun.

The Role: Computer Vision Engineer

What You’ll Get To Do:

We are looking for a production-oriented, computer vision engineer to design, develop and implement computer vision and machine learning cutting-edge technologies to ensure our market-leading position in sports player tracking and broadcast solutions.

• Design, prototype, implement and test software and computer vision & machine learning algorithms in Python and C++.
• Develop and optimize real-time and high-accuracy sports solutions with modern CV: object detection, recognition and tracking, camera calibration, 3D reconstruction, etc.
• Leverage Amazon Web Services (EC2 and S3) to run algorithms on a large number of servers in the cloud.
• Implement and provide best-practices for maintainable software development, including deployment process, documentation, and adherence to and improvement of coding standards.
• Support and monitor live systems, including on-call rotation for computer vision systems during sports seasons.
• Manage interdisciplinary projects in collaboration with different groups within the company.
• Continuously learn new applications and apply learnings to new challenges.

Our Software Stack:

• Languages for scientific computing: Python 3, modern C++
• Frameworks: OpenCV, CUDA, FFmpeg, Tensorflow, NumPy, scikit-learn
• Codebase: Git with GitLab
• Cloud services: AWS EC2 and S3
• Productivity: Slack, Notion, Google Apps

Requirements

What You Have:

• 2+ years of industrial experience in the full development life cycle: design, prototype, implementation, testing, and performance evaluation of software
• 2+ years of industrial experience in computer vision and modern machine learning algorithms, such as deep learning
• BSc or MSc in computer science or a related degree, with strong software engineering and modern computer vision components
• Solid experience in software engineering: modern software development practices and tools, concurrent and distributed programming, operating systems, computer networks, database systems
• Fully proficient in Python and modern C++ with exposure to functional & object-oriented programming paradigms. Other languages such as Rust will be considered as an advantage for some of our multidisciplinary projects.
• Experience with high-performance computing, GPU computing, and real-time systems would set you apart

Our Work Environment and What You Will Benefit From:

• Cutting-edge products to work on for major professional sports leagues and teams
• Team-oriented, start-up mindset, innovative and dynamic environment, which encourages growth and provides opportunities to make an impact
• Flat hierarchy and collaborative management led by highly-qualified technical leads