DOMAIN ADAPTATION FOR SEMANTIC SEGMENTATION VIA EXPLOITING WEAK LABELS

摘要

Systems and methods for adapting semantic segmentation across domains is provided. The method includes inputting (720) a source image into a segmentation network, and inputting (710) a target image into the segmentation network. The method further includes identifying (760) category wise features for the source image and the target image using category wise pooling, and discriminating (780) between the category wise features for the source image and the target image. The method further includes training (730) the segmentation network with a pixel-wise cross-entropy loss on the source image, and a weak image classification loss and an adversarial loss on the target image, and outputting a semantically segmented target image.