What are some deep learning models that can use supplementary information other than RGB channels for image segmentation?
For example, imagine a poorly shot image of a river (blue) that shows a gap, and the supplementary information is detailed flow directions (arrows), which helps to show the river's true shape (no gap in reality). To get the river shape, most image segmentation models I see, such as U-Net, only use RGB channels.
Are there any neural network models that can use this kind of auxiliary information along with RGB channels during training for the image segmentation task?
