Home

Awesome

D&D

This repo contains the code of our paper:

D&D: Learning Human Dynamics from Dynamic Camera
Jiefeng Li, Siyuan Bian, Chao Xu, Gang Liu, Gang Yu, Cewu Lu
ECCV 2022 Oral

<!-- [[`Paper`](https://openaccess.thecvf.com/content/CVPR2021/html/Li_HybrIK_A_Hybrid_Analytical-Neural_Inverse_Kinematics_Solution_for_3D_Human_CVPR_2021_paper.html)] [[`Supplementary Material`](https://openaccess.thecvf.com/content/CVPR2021/supplemental/Li_HybrIK_A_Hybrid_CVPR_2021_supplemental.zip)] [[`arXiv`](https://arxiv.org/abs/2011.14672)] [[`Project Page`](https://jeffli.site/HybrIK/)] -->

dnd

Train from scratch

python ./scripts/train.py --exp-id ${EXPID} --cfg ${CONFIG} --gpu ${GPU} --seed 1

Evaluation

python ./scripts/validate.py --cfg ${CONFIG} --ckpt ${CKPT} --gpu ${GPU}

Citing

If you find our code or paper useful, please consider citing

@inproceedings{li2022dnd,
    title={D\&D: Learning Human Dynamics from Dynamic Camera},
    author={Li, Jiefeng and Bian, Siyuan and Xu, Chao and Liu, Gang and Yu, Gang and Lu, Cewu},
    booktitle={European Conference on Computer Vision},
    year={2022}
}