Case Study: Nearmap achieves scalable machine learning and faster computer vision analytics with Amazon Web Services

Nearmap Uses Amazon SageMaker to Scale Machine Learning of Complex Computer Vision Analytics

nearmap needed a scalable way to train complex computer vision and machine learning models on more than 50 PB of aerial imagery. Using Amazon Web Services, specifically Amazon SageMaker and Amazon S3, the company moved away from time-consuming local data syncing and infrastructure-heavy model training.

Amazon Web Services implemented a streamlined workflow that lets nearmap train and deploy models directly from Amazon S3 on Amazon SageMaker, with parallel experiments and less operational overhead. The result was faster model training and deployment, virtually infinite scalable storage, and the ability to analyze urban tree cover across cities and over time—for example, showing Adelaide, Australia lost 9.8% of its relative tree cover over a decade.

nearmap

Nagita Mehr Seresht

Senior Director of AI Model Research and Development