Back to the Future Cyclopean Stereo: a human perception approach combining deep and geometric constraints
This incremental approach addresses stereo vision problems for applications in virtual reality and robotics by enhancing geometric modeling.
The paper tackles stereo vision by combining analytical 3D surface models with learned features to handle depth discontinuities and occlusions, achieving results on par with state-of-the-art data-driven methods while improving visual quality.
We innovate in stereo vision by explicitly providing analytical 3D surface models as viewed by a cyclopean eye model that incorporate depth discontinuities and occlusions. This geometrical foundation combined with learned stereo features allows our system to benefit from the strengths of both approaches. We also invoke a prior monocular model of surfaces to fill in occlusion regions or texture-less regions where data matching is not sufficient. Our results already are on par with the state-of-the-art purely data-driven methods and are of much better visual quality, emphasizing the importance of the 3D geometrical model to capture critical visual information. Such qualitative improvements may find applicability in virtual reality, for a better human experience, as well as in robotics, for reducing critical errors. Our approach aims to demonstrate that understanding and modeling geometrical properties of 3D surfaces is beneficial to computer vision research.