It was a semantic image segmentation challenge containing multi-class segmentations (road, buildings, vegetation,Green farms,Dry farms, barren land, water body and other) and our task was to extract the vegetation class, which had a pixel density (255, 201, 0, 255)
Approach Used : We used Different Pretrained Semantic Segmentation Models and Ensembled them to give the final prediction.
Challenge Link : https://www.kaggle.com/c/sematic-image-segmentation-challenge
Leaderboard Link: https://www.kaggle.com/c/sematic-image-segmentation-challenge/leaderboard