AI-Extracted Asian Building Footprints

submited by

Style Pass

2024-07-11 13:00:08

I recently came across a paper titled "Mapping 280 Million Buildings in East Asia Based on VHR Images". It was published in May and details an AI model that has extracted 281,093,433 building footprints from 2,897 cities across Eastern Asia.

The paper was written by eight authors affiliated with the University of Wisconsin, Guangdong Provincial Key Laboratory, Sun Yat-sen University and Wuhan University. 100 TB of 50cm, zoom-level 18 source imagery from Google Earth taken between 2020 and 2022 was used for this exercise. The paper states the average overall accuracy is 89.63% and has an F1 score of 82.55%. Eight pieces of prior art focused on deep learning approaches were studied prior to this model's development.

In this post, I'll examine the resulting comprehensive large-scale mapping framework's "CLSM" building footprint dataset.

I'm using a 6 GHz Intel Core i9-14900K CPU. It has 8 performance cores and 16 efficiency cores with a total of 32 threads and 32 MB of L2 cache. It has a liquid cooler attached and is housed in a spacious, full-sized, Cooler Master HAF 700 computer case. I've come across videos on YouTube where people have managed to overclock the i9-14900KF to 9.1 GHz.