arthur151 / relative_human

Relative Human dataset, CVPR 2022

Relative Human (RH) contains multi-person in-the-wild RGB images with rich human annotations, including:

• Depth layers (DLs): relative depth relationship/ordering between all people in the image.
• Age group classfication: adults, teenagers, kids, babies.
• Others: Genders, Bounding box, 2D pose.

RH is introduced in CVPR 2022 paper Putting People in their Place: Monocular Regression of 3D People in Depth.

[Project Page] [Video] [BEV Code]

Download

[Google drive]
[Baidu drive]

Leaderboard

See Leaderboard.

Why do we need RH?

Existing 3D datasets are poor in diversity of age and multi-person scenories. In contrast, RH contains richer subjects with explicit age annotations in the wild. We hope that RH can promote relative research, such as monocular depth reasoning, baby / child pose estimation, and so on.

How to use it?

We provide a toolbox for data loading, visualization, and evaluation.

To run the demo code, please download the data and set the dataset_dir in demo code.

To use it for training, please refer to BEV for details.

Re-implementation

To re-implement RH results (in Tab. 1 of BEV paper), please first download the predictions from here, then

cd Relative_Human/
# BEV / ROMP / CRMH : set the path of downloaded results (.npz) in RH_evaluation/evaluation.py, then run
python -m RH_evaluation.evaluation

cd RH_evaluation/
# 3DMPPE: set the paths in eval_3DMPPE_RH_results.py and then run
python eval_3DMPPE_RH_results.py
# SMAP: set the paths in eval_SMAP_RH_results.py and then run
python eval_SMAP_RH_results.py

Citation

Please cite our paper if you use RH in your research.

@InProceedings{sun2022BEV,
author = {Sun, Yu and Liu, Wu and Bao, Qian and Fu, Yili and Mei, Tao and Black, Michael J},
title = {Putting People in their Place: Monocular Regression of {3D} People in Depth}, 
booktitle = {IEEE/CVF Conf.~on Computer Vision and Pattern Recognition (CVPR)}, 
year = {2022}
}

relative_human's Issues

Trans

感谢您的分享！有一个问题想向您请教。
在测试中直接用到了ROMP结果的trans，ROMP得到的相机参数是弱透视投影下的（s,tx,ty），他是如何转换成trans的呢？会提供相应的代码吗？

Annotation quality issues

Thanks for the effort to create this dataset!

When inspecting the annotations, I found some quality issues with the annotations. For some images, the annotations do not seem to be exhaustive. Some clearly visible persons in the foreground are missing in the annotations, such as the one the left in the following image (105520.jpg):

There are also cases where the full box annotation is only covering part of the person, even if other parts are clearly visible, such the old man riding a horse in this image (100134.jpg):

There also seem to be overlap between the training set and the eval/test set. For example, 109136.jpg in the validation set appears to be a resized version of 000154.jpg in the training set:

Would the authors mind looking into these issues? Thanks!