Hi, Is there any chance for your to share the trained model and example data to try the results? The paper and code looks promising, but a bit hard to reproduce it due to data access from public dataset. Thanks!