Skip to main content

Table 3 Per class average accuracy for HMDB51

From: Action recognition using length-variable edge trajectory and spatio-temporal motion skeleton descriptor

brushhair:0.89

eat:0.53

kiss:0.80

shakehands:0.62

swordexercise:0.19

cartwheel:0.63

fallfloor:0.53

laugh:0.69

football:0.76

swordfight:0.29

catch:0.67

fencing:0.64

pick:0.27

shootbow:0.78

talk:0.64

chew:0.71

backhandflip:0.76

pour:0.77

shotgun:0.48

throw:0.04

claphands:0.77

golf:0.97

pullup:0.97

sit:0.82

turn:0.62

climb:0.64

handstand:0.66

punch:0.33

situp:0.73

walk:0.57

climbstairs:0.68

hit:0.28

push:0.57

smile:0.19

wave:0.14

dive:0.57

hug:0.83

pushup:0.82

smoke:0.50

 

drawsword:0.46

jump:0.44

ridebike:0.72

somersault:0.52

 

dribble:0.88

kickball:0.49

ridehorse:0.69

standup:0.67

 

drink:0.67

kick:0.27

run:0.49

swingbaseball:0.06

Â