Skip to content

Latest commit

 

History

History
29 lines (26 loc) · 1.54 KB

File metadata and controls

29 lines (26 loc) · 1.54 KB

2341-People-Gesture-Recognition-Data-in-Online-Conference-Scenes

Description

2,341 People Gesture Recognition Data in Meeting Scenes includes Asians, Caucasians, blacks, and browns, and the age is mainly young and middle-aged. It collects a variety of indoor office scenes, covering meeting rooms, coffee shops, libraries, bedrooms, etc. Each person collected 18 pictures and 2 videos. The pictures included 18 gestures such as clenching a fist with one hand and heart-to-heart with one hand, and the video included gestures such as clapping. For more details, please refer to the link: https://www.nexdata.ai/datasets/computervision/1292?source=Github

Data size

2,341 people, each person collects 2 videos and 18 images

Race distribution

786 Asians, 1,002 Caucasians, 401 black, 152 brown people

Gender distribution

1,209 males, 1,132 females

Age distribution

from teenagers to the elderly, mainly young and middle-aged

Collection environment

indoor office scenes, such as meeting rooms, coffee shops, libraries, bedrooms, etc.

Collection diversity

different gestures data, different races, different age groups, different meeting scenes

Collection equipment

cellphone, using the cellphone to simulate the perspective of laptop camera in online conference scenes

Collection content

collecting the gestures data in online conference scenes

Data format

.mp4, .mov, .jpg

Accuracy rate

the accuracy exceeds 97% based on the accuracy of the actions; the accuracy of action naming is more than 97%

Licensing Information

Commercial License