Robotcar Dataset

This is the code to generate the corresponding ground truth depth map of the images in Robotcar dataset,

and the main doc is "eval_oxford.py".

This test sequence "test_files" follows the paper:

@inproceedings{vankadari2020unsupervised,
  title={Unsupervised monocular depth estimation for night-time images using adversarial domain feature adaptation},
  author={Vankadari, Madhu and Garg, Sourav and Majumder, Anima and Kumar, Swagat and Behera, Ardhendu},
  booktitle={European Conference on Computer Vision},
  pages={443--459},
  year={2020},
  organization={Springer}
}
@article{zhao2022unsupervised,
  title={Unsupervised monocular depth estimation in highly complex environments},
  author={Zhao, Chaoqiang and Tang, Yang and Sun, Qiyu},
  journal={IEEE Transactions on Emerging Topics in Computational Intelligence},
  volume={6},
  number={5},
  pages={1237--1246},
  year={2022},
  publisher={IEEE}
}

⚙️ Acknowledgement

The basic code and test sequence are provided by the authors of "Unsupervised monocular depth estimation for night-time images using adversarial domain feature adaptation".