This repo contains the code from the benchmark blog post for the Open AI Caribbean Challenge: Mapping Disaster Risk from Aerial Imagery competition sponsored by MathWorks.
BenchmarkBlog.mlx is the MATLAB live script of the benchmark code with detailed instructions on how to load, process, and predict on the data. resnet_presaved.mat contains the pretrained ResNet-18 model used in the benchmark.
Note: To access these files you have to download or clone the repo.
Benchmark_code.m is the MATLAB code file of the benchmark, which is viewable directly on GitHub.