Micro Expression Recognition via Dual-Stream Spatiotemporal Attention Network

Figure 6

The performance of DSTAN for each subject: (a) SMIE, (b) CASME, and (c) CASME II.

(a)

(b)

(c)