Shape from Polarization for Complex Scenes in the Wild
This work addresses the challenge of scene-level shape from polarization for applications in computer vision and robotics, though it is incremental as it builds on existing SfP methods with new data and model improvements.
The authors tackled the problem of estimating surface normals from a single polarization image for complex real-world scenes, a task previously limited to single objects, by introducing the first real-world scene-level dataset and a learning-based framework that significantly outperforms existing models on two datasets.
We present a new data-driven approach with physics-based priors to scene-level normal estimation from a single polarization image. Existing shape from polarization (SfP) works mainly focus on estimating the normal of a single object rather than complex scenes in the wild. A key barrier to high-quality scene-level SfP is the lack of real-world SfP data in complex scenes. Hence, we contribute the first real-world scene-level SfP dataset with paired input polarization images and ground-truth normal maps. Then we propose a learning-based framework with a multi-head self-attention module and viewing encoding, which is designed to handle increasing polarization ambiguities caused by complex materials and non-orthographic projection in scene-level SfP. Our trained model can be generalized to far-field outdoor scenes as the relationship between polarized light and surface normals is not affected by distance. Experimental results demonstrate that our approach significantly outperforms existing SfP models on two datasets. Our dataset and source code will be publicly available at https://github.com/ChenyangLEI/sfp-wild