Domain Adaptation through Photorealistic Enhanced Images for Semantic Segmentation

Figure 1

Ground truth and photorealistic enhanced images. (a) Annotation data. (b) Input image. (c) photorealistic enhanced image.