CVPR 2024
- [April 2024] Inference scripts for video or image to depth.
- [March 2024] Pretrained checkpoints for NYUv2 and KITTI datasets.
- [March 2024] Training and Evaluation code released!
- [Feb 2024] ECoDepth accepted in CVPR'2024.
git clone https://github.com/Aradhye2002/EcoDepth
cd EcoDepth
conda env create -f env.yml
conda activate ecodepthYou can take a look at the dataset preparation guide for NYUv2 and KITTI from here. After downloading the datasets, change the data paths in the respective bash files to point to your dataset location where you have downloaded the datasets. Alternatively, you can also make a symbolic link of the dataset folders like so:
cd depth
mkdir data
cd data
ln -s <path_to_kitti_dataset> kitti
ln -s <path_to_nyu_dataset> nyuNote the dataset structure inside the path you have given in the bash files should look like this:
NYUv2:
nyu
├── nyu_depth_v2
│ ├── official_splits
│ └── syncKITTI:
kitti
├── KITTI
│ ├── 2011_09_26
│ ├── 2011_09_28
│ ├── 2011_09_29
│ ├── 2011_09_30
│ └── 2011_10_03
└── kitti_gt
├── 2011_09_26
├── 2011_09_28
├── 2011_09_29
├── 2011_09_30
└── 2011_10_03Please download the pretrained weights from this link and save .ckpt weights inside <repo root>/depth/checkpoints directory.
Also download the v1-5 checkpoint of stable-diffusion and put it in the <repo root>/checkpoints directory. Please create an empty directory if you find that such a path does not exist. Note that this checkpoints folder is different from the one above.
To perform inference on any RGB image or video use the infer_{outdoor,indoor}.sh file. Set the --img_path argument to the image you would to get the depth for and the --video_path to the video from which to produce the depth. In case you only wish to infer on an img or video, simply remove the other argument. Then enter the depth directory by executing cd depth and run:
-
Infer on outdoor scenes:
bash infer_outdoor.sh -
Infer on outdoor scenes:
bash infer_indoor.sh
To evaluate the model performance on NYUv2 and KITTI datasets, use the test_{kitti, nyu} file. The trained models are publicly available, download the models using above links. Then, navigate to the depth directory and follow the instructions outlined below:
-
Evaluate on NYUv2 dataset:
bash test_nyu.sh <path_to_saved_model_of_NYU> -
Evaluate on KITTI dataset:
bash test_kitti.sh <path_to_saved_model_of_KITTI>
We trained our models on 32 batch size using 8xNVIDIA A100 GPUs. Inside the train_{kitti,nyu}.sh set the NPROC_PER_NODE variable and --batch_size argument to the desired values as per your system resources. For our method we set them as NPROC_PER_NODE=8 and --batch_size=4 (resulting in a total batch size of 32). Afterwards, navigate to the depth directory by executing cd depth and follow the instructions:
-
Train on NYUv2 dataset:
bash train_nyu.sh -
Train on KITTI dataset:
bash train_kitti.sh
If you have any questions about our code or paper, kindly raise an issue on this repository.
We thank Kartik Anand for assistance with the experiments. Our source code is inspired from VPD and PixelFormer. We thank their authors for publicly releasing the code.
If you find our work useful in your research, please consider citing using:
@InProceedings{Patni_2024_CVPR,
author = {Patni, Suraj and Agarwal, Aradhye and Arora, Chetan},
title = {ECoDepth: Effective Conditioning of Diffusion Models for Monocular Depth Estimation},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2024},
pages = {28285-28295}
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent