400 masks: 150 plaster masks; 100 resin masks; 100 silicone masks, 50 silicone head models Each mask includes 252 videos under various lighting and with various devices (18 lighting environments, 14 devices, 18X14=252) This database only contains attack videos, no real person videos

100 real people 170 half-face masks 110 paper masks 30 cloth masks 5 transparent masks 95 low-precision online purchased masks 300 3D printed masks (150 18cm masks, 150 15cm masks) Each mask or real person is captured with 2 devices under 4 lighting conditions, totaling 8 videos per mask/person, each video about 60 seconds.

1 person + 7 videos: 1. Speaking facing the camera 2. Walking forward and backward facing the camera 3. Walking along the diagonal 4. Walking facing the camera 5. Walking with mask facing the camera 6. Walking left and right facing the camera 7. Head rotation age distribution: 18-:393, 18-35:2781, 35-55:777, 55+:650

13 videos per person, acquired with 3DMD camera 1 person 13 expressions of high-definition 3D face data

10 videos per person, acquired with 3DMD. Collected individuals are consistent with the 8400 multi-racial data (5000+ Chinese, 3000+ other nationalities)

1 person 12 videos: 1. 4 videos of human body front, 45 degrees left, 45 degrees right, and back, captured 3 sets using RealSense D455 2. Each person's height, weight, body fat, waist circumference, chest circumference, hip circumference, etc. data measured twice, gender, BMI meet certain distribution

1 person + 48 videos 1. 24 videos before makeup and 24 videos after makeup 2. Before and after makeup videos include 9 multi-pose multi-expression videos captured by mobile phone and RealSense SR300 respectively, and 6 videos captured by Hikvision surveillance camera

99000 registration videos + 1360 videos in different scenes 1. Single person registration video: One person under 6 cameras, 5 actions, 2 states of wearing/removing glasses, 60 videos per person, total 1650 people, 99000 videos. 2. 1650 people randomly grouped into 80 groups with 15-25 people per group. Each group is filmed for 15-20 minutes in the following 4 scenes, with one video collected by each camera to obtain data. Multi-person different scene videos: 7 cameras in cafe scene, 3 cameras in elevator scene, 4 cameras in escalator scene, 3 cameras in L-shaped corridor scene, 80 groups, total 1360 videos. age distribution: 18-:14, 18-35:1427, 35-55:223, 55+:4

nan

1. Single person registration video: Each person collects three videos: wearing a mask, half-wearing a mask, and not wearing a mask. 2. 300 people are randomly grouped into about 20 groups of 10-20 people. Each group is filmed for 5-10 minutes in the following 2 scenes, with each camera collecting 3 videos, for wearing a mask, half-wearing a mask, and not wearing a mask, respectively. Multi-person different scene videos: 6 cameras in entrance scene, 2 cameras in indoor dim light scene.