Skip to content

Nexdata-AI/18880-Images-of-466-People-3D-Instance-Segmentation-and-22-Landmarks-Annotation-Data-of-Human-Body

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 

Repository files navigation

18880-Images-of-466-People-3D-Instance-Segmentation-and-22-Landmarks-Annotation-Data-of-Human-Body

Description

18,880 Images of 466 People - 3D Instance Segmentation and 22 Landmarks Annotation Data of Human Body. The dataset diversity includes multiple scenes, light conditions, ages, shooting angles, and poses. In terms of annotation, we adpoted instance segmentation annotations on human body. 22 landmarks were also annotated for each human body. The dataset can be used for tasks such as human body instance segmentation and human behavior recognition.

For more details, please refer to the link: https://www.nexdata.ai/datasets/computervision/1040?source=Github

Data size

466 people, the number of images is 18,880

Race distribution

Asian

Gender distribution

233 males, 233 females

Age distribution

299 children and teenagers, 167 adults

Collecting environment

including indoor scenes and outdoor scenes

Data diversity

multiple scenes, light conditions, ages, shooting angles, and poses

Device

RealSense Depth Camera D435i

Data format

the image data format is .jpg and .png, the annotation file format are .json,the file format of camera parameters is .txt

Collecting content

3D images of various human body poses ((the rgb channel and depth channel have been registered))

Annotation content

instance segmentation annotation of human body , 22 landmarks annotation of human body

Accuracy

Accuracy requirement: the mask edge location errors in x and y directions are less than 3 pixels, which is considered as a qualified annotation; Accuracy requirement of segmentation annotation: the annotation part (each part of mask) is regarded as the unit, the accuracy rate shall be more than 95%; Accuracy requirement of landmark annotation: the annotation part (each landmark) is regarded as the unit, the accuracy rate shall be more than 95%; Annotating accuracy = number of correct annotations / total number of annotations

Licensing Information

Commercial License