UIERL: Internal-External Representation Learning Network for Underwater Image Enhancement

Recently, many underwater image enhancement (UIE) methods have been proposed. Although much progress has been made, they still face two issues: (1) There exists a significant region-wise quality difference in a single underwater image due to the underwater imaging process, especially in regions with different scene depths. However, existing methods neglect this internal characteristic, yielding inferior performance; (2) Due to the unique acquisition approach, underwater image acquisition tools usually capture multiple images in the same or similar scenes. The underwater images to be enhanced in practical usage are highly correlated. However, when processing a single image, existing methods ignore the rich external information provided by related images. There is still room for improvement in their performance. This paper proposes an internal-external representation learning (UIERL) network to better perform UIE tasks with internal and external information, simultaneously. In the internal representation learning stage, a depth-based region feature guidance network is designed, including a region segmentation based on scene depth to sense regions with different quality levels, and a region-wise space encoder for region-specific feature learning, which provides an effective guidance for global features to guide intra-image differentiated enhancement. In the external representation learning stage, we first propose an external information extraction network to mine rich external information from related images. Then, internal and external features interact with each other via the proposed external-assist-internal module and internal-assist-external module, fully exploiting the rich internal and external information to better enhance a single image. Extensive experiments demonstrate the superiority of our UIERL.