Polarized Reflection Removal with Perfect Alignment in the Wild
This work improves reflection removal for computer vision applications, but it is incremental as it builds on existing methods with dataset and loss enhancements.
The paper tackles the problem of removing reflections from polarized images in the wild by addressing dataset misalignment issues and proposing a two-stage model with a novel perceptual NCC loss, achieving state-of-the-art performance on reflection removal.
We present a novel formulation to removing reflection from polarized images in the wild. We first identify the misalignment issues of existing reflection removal datasets where the collected reflection-free images are not perfectly aligned with input mixed images due to glass refraction. Then we build a new dataset with more than 100 types of glass in which obtained transmission images are perfectly aligned with input mixed images. Second, capitalizing on the special relationship between reflection and polarized light, we propose a polarized reflection removal model with a two-stage architecture. In addition, we design a novel perceptual NCC loss that can improve the performance of reflection removal and general image decomposition tasks. We conduct extensive experiments, and results suggest that our model outperforms state-of-the-art methods on reflection removal.