Virtual View Networks for Object Reconstruction
This extends rigid structure-from-motion techniques to challenging settings like single-image reconstruction, benefiting computer vision applications.
The paper tackles object reconstruction from limited images (single or few far-apart views) by synthesizing virtual views using geodesics on networks connecting objects with similar viewpoints, achieving accurate shape reconstruction on challenging PASCAL VOC data.
All that structure from motion algorithms "see" are sets of 2D points. We show that these impoverished views of the world can be faked for the purpose of reconstructing objects in challenging settings, such as from a single image, or from a few ones far apart, by recognizing the object and getting help from a collection of images of other objects from the same class. We synthesize virtual views by computing geodesics on novel networks connecting objects with similar viewpoints, and introduce techniques to increase the specificity and robustness of factorization-based object reconstruction in this setting. We report accurate object shape reconstruction from a single image on challenging PASCAL VOC data, which suggests that the current domain of applications of rigid structure-from-motion techniques may be significantly extended.