Data

The basic principle behind our approach is to obtain dense and accurate 3D reconstructions from large collections of images with off-the-shelf Structure from Motion (SfM), using Colmap. This provides us (pseudo-)ground truth poses and densified depth maps. We then subsample these image collections into smaller, random subsets (e.g. pairs of images for stereo, or 5 to 25 images for multiview reconstruction with SfM) and evaluate different methods against the "ground-truth".

We publish training data with images, poses, depth maps, and co-visibility estimates. We also provide a validation set in the format expected by the benchmark, to allow challenge participants to tune their methods before submission. The test set will remain private: we have published a subset of 100 images per scene, which participants can use to extract features and, optionally, matches. Please refer to the submission page for details.

The Phototourism dataset

• Download link

In order to learn and evaluate models that can perform well under a wide range of situations, it is of paramount importance to collect information from multiple sensors obtained at different times, from different viewpoints, and with occlusions. A natural solution is thus to turn to photo-tourism data. In this dataset we rely on 26 photo-tourism image collections of popular landmarks originally collected by the Yahoo Flickr Creative Commons 100M (YFCC) dataset and Reconstructing the world in six days. They range from ~100s to ~1000s of images.

We provide the full data, with ground truth, for 16 scenes, which can be optionally used for training. We format 3 of them (Reichstag, Sacre Coeur and Saint Peter's Square) for validation, and reserve 10 for testing.

We provide examples to parse the training data on the benchmark repository: please refer to this notebook for details.