README.md

This is offical codes TripleDNet: Exploring Depth Estimation with Self-Supervised Representation Learning

If you find our work useful in your research please consider citing our paper:

@inproceedings{triplednetdepth,
  title={TripleDNet: Exploring Depth Estimation with Self-Supervised Representation Learning},
  author={Senturk, Ufuk Umut and Akar, Arif and Ikizler-Cinbis, Nazli},
  booktitle={BMVC},
  year={2022}
}

Models

Coming soon!

Setup

Requirements:

• PyTorch1.1+, Python3.5+, Cuda10.0+
• mmcv==0.4.4

conda create --name featdepth python=3.7
conda activate featdepth

# this installs the right pip and dependencies for the fresh python
conda install ipython
conda install pip

# install required packages from requirements.txt
pip install -r requirements.txt

KITTI training data

Our training data is the same with other self-supervised monocular depth estimation methods, please refer to monodepth2 to prepare the training data.

API

We provide an API interface for you to predict depth and pose from an image sequence and visulize some results. They are stored in folder 'scripts'.

eval_depth.py is used to obtain kitti depth evaluation results.

infer.py or infer_singleimage.py is used to generate depth maps from given models.

Training

You can use following command to launch distributed learning of our model:

/path/to/python -m torch.distributed.launch --master_port=9900 --nproc_per_node=1 train.py --config /path/to/cfg_kitti_tripleD.py --work_dir /dir/for/saving/weights/and/logs'

or undistributed learning

/path/to/python -m --master_port=9900 --config /path/to/cfg_kitti_tripleD.py --work_dir /dir/for/saving/weights/and/logs'

Here nproc_per_node refers to GPU number you want to use, which is 4 in our case.

Acknowledgement

This repository is based on FeatDepth. We thank authors for their contributions.