PLADE-net

Repository of our CVPR2021 paper "PLADE-Net: Towards Pixel-Level Accuracy for Self-Supervised Single-View Depth Estimation With Neural Positional Encoding and Distilled Matting Loss" Paper and supplemental materials:https://openaccess.thecvf.com/content/CVPR2021/html/Gonzalez_PLADE-Net_Towards_Pixel-Level_Accuracy_for_Self-Supervised_Single-View_Depth_Estimation_With_CVPR_2021_paper.html

Pre-trained models (comming soon)

Computed inverse depth maps (comming soon)

Cite our paper:

@InProceedings{Gonzalez_2021_CVPR,
    author    = {Gonzalez, Juan Luis and Kim, Munchurl},
    title     = {PLADE-Net: Towards Pixel-Level Accuracy for Self-Supervised Single-View Depth Estimation With Neural Positional Encoding and Distilled Matting Loss},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2021},
    pages     = {6851-6860}
}

New: Added files to train our PLADE-net on KITTI and KITTI + CityScapes

Coming soon: Code cleaning :)

Not for commercial use If wanted for commercial use contact juanluisgb@kaist.ac.kr