Open 51616 opened 5 years ago
@51616 You can find everything we knew on the competition overview page https://www.crowdai.org/challenges/mapping-challenge.
This dataset was quite simple and not very diverse, which means that it the model we trained may work incredibly well on similar images, but in order to get good generalization, you need to construct a new dataset from more diverse data sources.
What is the specification of the training data (e.g. spatial resolution or the satellite that took the images)? Also, the training data seems like it is blue shifted compared to other dataset that I have and the pretrained model has poor performance in that dataset. Can you provide the data specification so that I can modify my dataset to be closer to this one. Thanks! :)