Research on a Microexpression Recognition Technology Based on Multimodal Fusion

Figure 7

Feature vectors (b, d) obtained from the text data contained in images (a) and (c) after CNN + LSTM neural network.

(a)

(b)

(c)

(d)