1. Overview

The THU-Bi-Hand binocular hand pose dataset contains 446620 frames of binocular images captured from 10 subjects. The images are captured by Leap Motion, under the resolution of 640x480 for both left and right images. 3D positions of the five fingertips and the wrist are annotated using 3D Guidance trakSTAR.

Each subject was asked to perform the 16 kinds of basic hand poses in the order as shown in Fig. 1, with both translation and rotation of the hand. The transforming poses between each pair of the adjacent basic poses were also captured, and the subject was asked to perform the transforming procedure several times before he performing the next basic pose. Similarly, translation and rotation of the hand were also involved during the transforming procedure. After all basic poses were performed, another similar procedure was carried out but in the reverse order of the 16 basic poses. Afterwards, the subject was asked to do transformations between several pairs of basic poses that were not adjacent, for example, changing between the second basic pose and the fourth basic pose, the second pose and the fifth pose, etc. The pairs of poses were chosen randomly by the subject himself.

During the sampling procedure, the subject could move his fingers freely, like swinging and clicking. Translations along all three directions were allowed as long as the hand did not disappear from the valid imaging area. Rotations within 90 degrees around all the three axis were allowed, while the initial hand was in the plane parallel to the imaging plane of the Leap Motion and extended perpendicular to the baseline connecting the left camera and the right camera of the Leap Motion.

The training set contains totally 356522 samples, including all the samples of seven subjects and half of the samples of another two subjects. The test set contains the remaining 90098 samples (including half of the samples of two subjects and all the samples of the remaining one subject).

2. Citing the dataset

@article{wang2020bi,

title={Bi-Stream Pose-Guided Region Ensemble Network for Fingertip Localization From Stereo Images},

author={Wang, Guijin and Zhang, Cairong and Chen, Xinghao and Ji, Xiangyang and Xue, Jing-Hao and Wang, Hang},

journal={IEEE Transactions on Neural Networks and Learning Systems},

year={2020},

publisher={IEEE}

}

3. Download

Please contact Cairong Zhang: zcr17@mails.tsinghua.edu.cn.

4. Publication

Bi-Stream Pose-Guided Region Ensemble Network for Fingertip Localization From Stereo Images

Guijin Wang ; Cairong Zhang ; Xinghao Chen ; Xiangyang Ji ; Jing-Hao Xue ; Hang Wang

IEEE Transactions on Neural Networks and Learning Systems ( TNNLS, Early Access ) 2020

5. Contact

For questions about the dataset, contact Cairong Zhang: zcr17@mails.tsinghua.edu.cn.