Skip to main content
. 2019 Dec 9;19(24):5429. doi: 10.3390/s19245429

Figure 5.

Figure 5

The proposed multimodal model architecture of 3DCNN + LSTM: (a) Early fusion multimodal; (b) Middle fusion multimodal; (c) Late fusion multimodal.