Learning to Jointly Generate and Separate Reflections

Existing learning-based single image reflection removal methods using paired training data have limitations about the generalization capability of dealing with real-world reflections due to the limited variations in training pairs. In this work, we propose to jointly generate and separate reflections within a weakly-supervised learning framework, aiming to model the reflection image formation more comprehensively with abundant unpaired supervision. By imposing the entanglement and disentanglement mechanisms, the proposed framework elegantly integrates two independent stages of reflection generation and separation into a unified model. For better performance, the image gradient constraint is incorporated into the concurrent training process of the multi-task learning as well. In particular, we built up an unpaired reflection dataset with 4,027 images, which is useful for investigating the problem of reflection removal in the weakly supervised learning manner, and further improving model performance. Extensive experiments on a public benchmark dataset show that our framework performs favorably against state-of-the-art methods and consistently produces visually appealing results.