A Multimodal Information Fusion Model for Robot Action Recognition with Time Series

Table 3

Accuracy of action recognition obtained by different methods.

Dataset\Method	RNN	LSTM	Reference [28]	Reference [29]	Reference [30]	Proposed
NTU-RGB+D	0.8284 ± 0.0234	0.8303 ± 0.0132	0.8389 ± 0.0207	0.8521 ± 0.0265	0.8612 ± 0.0201	0.8683 ± 0.0215
MSR action 3D	0.9297 ± 0.0354	0.9265 ± 0.0268	0.9432 ± 0.0193	0.9542 ± 0.0213	0.9460 ± 0.0321	0.9529 ± 0.0344