DELTAR: Depth Estimation from a Light-weight ToF Sensor And RGB Image

Project Page | Paper

DELTAR: Depth Estimation from a Light-weight ToF Sensor And RGB Image
Yijin Li, Xinyang Liu, Wenqi Dong, Han Zhou, Hujun Bao, Guofeng Zhang, Yinda Zhang, Zhaopeng Cui
ECCV 2022

News: You can find the code of calibration here

Download Link

We provide the download link [google drive, baidu(code: 1i11)] to

• pretrained model trained on NYU.
• ZJUL5 dataset.
• demo data.

Run DELTAR

Installation

conda create --name deltar --file requirements.txt

Prepare the data and pretrained model

Download from the above link, and place the data and model as below:

deltar
├── data
│   ├── demo
│   └── ZJUL5
└── weights
    └── nyu.pt

Evaluate on ZJUL5 dataset

python evaluate.py configs/test_zjuL5.txt

Run the demo

python evaluate.py configs/test_demo.txt
python scripts/make_gif.py --data_folder data/demo/room --pred_folder tmp/room

Citation

If you find this code useful for your research, please use the following BibTeX entry.

@article{deltar,
  title={DELTAR: Depth Estimation from a Light-weight ToF Sensor and RGB Image},
  author={Li Yijin and Liu Xinyang and Dong Wenqi and Zhou han and Bao Hujun and Zhang Guofeng and Zhang Yinda and Cui Zhaopeng},
  booktitle={European Conference on Computer Vision (ECCV)},
  year={2022}
}

Acknowledgements

We would like to thank the authors of Adabins, LoFTR and Twins for open-sourcing their projects.