We introduce a camera relocalization pipeline that combines absolute pose regression (APR) and direct feature matching. By incorporating exposure-adaptive novel view synthesis, our method successfully addresses photometric distortions in outdoor environments existing photometric-based methods fail to handle. With domain-invariant matching, solution improves accuracy using semi-supervised learni...

By far, many stereo-matching techniques have been successfully proposed and applied in digital aerial photogrammetry. However, due to some problems such as large parallaxes, occlusions, geometric deformations, and repetitive patterns in convergent close range images, these methods may not be applicable to the same level of success as that of aerial imagery. In order to overcome these shortcomin...