Counted all of the yurts in Mongolia using machine learning

A developer trained a YOLO machine learning model to count yurts across Mongolia using satellite imagery from Google Maps. The project involved manually labeling over 10,000 yurts as training data, optimizing the search area using geographic data, and deploying the model across 120 Docker containers using Docker Swarm. The final count revealed 172,689 yurts throughout the country, with most concentrated in urban ger districts around Ulaanbaatar where rural migrants have settled

Computer Science and Programming

Counted all of the yurts in Mongolia using machine learning

A developer trained a YOLO machine learning model to count yurts across Mongolia using satellite imagery from Google Maps. The project involved manually labeling over 10,000 yurts as training data, optimizing the search area using geographic data, and deploying the model across 120 Docker containers using Docker Swarm. The final count revealed 172,689 yurts throughout the country, with most concentrated in urban ger districts around Ulaanbaatar where rural migrants have settled

@computer_science_and_programming

👍6

hottg.com/computer_science_and_programming/1598

9.79K viewsJun 19 at 06:10