Dataset Open Access
This dataset contains the synthetic and real data used in the article " Virtual Training for a Real Application: Accurate Object-Robot Relative Localization Without Calibration " to train 3 convolutional neural networks (CNNs) in order to perform uncalibrated relative localization of a cuboid block with respect to a robot, and to evaluate them.
It consists of a dataset composed of synthetic pictures for training the CNNs and a dataset of real pictures for evaluation.
The "synthetic" dataset is composed 3 sub-datasets (each of them composed of thousands of synthetic pictures and corresponding groundtruth) for training :
These sub-datasets are composed of raw data as well as post-treated data ready to be used for training CNNs, in CSV format and in Torch format (.t7).
The "real" dataset is composed of real pictures with a precisely localized cuboid block for evaluation only : UnLoc_real (~2.5 GB when extracted).
More information available on the project page : http://imagine.enpc.fr/~loingvi/unloc/
Loing et al. (2018). Dataset associated to arXiv:1902.02711