Case Study: Texas A&M University-Corpus Christi improves coastal shoreline classification with Google for Education's AutoML Vision

Texas A&M University - Corpus Christi improves shoreline classification accuracy to 91% with Google for Education

Texas A&M University - Corpus Christi's Harte Research Institute needed a way to automatically classify nearly nine thousand miles of Texas coastline from aerial imagery to create detailed Environmental Sensitivity Index maps. This manual process was slow and required expert scrutiny. To address this challenge, they turned to Google for Education, using Google Cloud's AutoML Vision to build a custom image classification model.

The solution from Google for Education involved training AutoML Vision on both single and multi-labeled datasets of aerial shoreline images. Implementing the multi-label classification feature significantly improved the model's accuracy, achieving a 0.952 average prediction rate with 91% precision and 90% recall. This allowed non-experts to assign sensitivity values, automating the process and making it much easier to build accurate, custom image classification models on their own data.

Texas A&M University - Corpus Christi