Detector-in-Detector: Multi-Level Analysis for Human-Parts (ACCV'2018)
By Xiaojie Li, Lu Yang, Qing Song, Fuqiang Zhou
This project provides the Human-Parts dataset used in DID-Net.
Vision-based person, hand or face detection approaches have achieved incredible success in recent years with the development of deep convolutional neural network (CNN). We take the inherent correlation between the body and body parts into account and propose a new framework to boost up the detection performance of the multi-level objects. In particular, we adopt region-based object detection structure with two carefully designed detectors to separately pay attention to the human body and body parts in a coarse-to-fine manner, which we call Detector-in-Detector network (DID-Net). The framework is trained in an end-to-end way by optimizing a multi-task loss. Due to the lack of human body, face and hand detection dataset, we have collected and labeled a new large dataset named Human-Parts with 14,962 images and 106,879 annotations. Experiments show that our method can achieve excellent performance on Human-Parts. A detailed introduction of Human-Parts can be found in the paper.
We hope that the Hunam-Parts dataset could also be used in other related tasks and applications.
Citation
If you find Human-Parts dataset or DID-Net useful in your research, please cite:
@article{didnet,
title={Detector-in-Detector: Multi-Level Analysis for Human-Parts},
author={Xiaojie Li, Lu yang, Qing Song, Fuqiang Zhou},
journal={arXiv preprint arXiv:****},
year={2019}
}
Data Downloads
BaiduDrive Link: https://pan.baidu.com/s/1bFDttum-6v1qbr23mpaNaw
DropBox Link: https://www.dropbox.com/s/3xxi4b7grx83hxi/Priv_personpart.tar?dl=0
Google Drive Link: https://drive.google.com/file/d/1L7oxFqRi63APVi-ffeK3L7dF_qTkZmbW/view?usp=sharing
Contact
Questions can also be left as issues in the repository.