Efficient Interaction Recognition through Positive Action Representation

Figure 8

Confusion matrix for different features for two approaches. (a)–(c) are the recognition results based on Positive Action representation; (d)–(f) are the results based on old one referred to [5].

(a) Joint features (new)

(b) Plane features (new)

(c) Velocity features (new)

(d) Joint features (old)

(e) Plane features (old)

(f) Velocity features (old)