gymprathap/Semantic-Segmentation-Aerial-Imagery-Dataset

The dataset comprises aerial imagery of Dubai acquired by MBRSC satellites and annotated with pixel-level semantic segmentation across 6 distinct classes. The dataset comprises a total of 72 images, which are organised into 6 larger tiles. The categories are as follows:

Credit: Humans in the Loop is releasing an openly accessible dataset that has been annotated for a collaborative project with the Mohammed Bin Rashid Space Centre in Dubai, United Arab Emirates.

Deep Learning Projects for Final Year