Two-Way Neural Network Chinese-English Machine Translation Model Fused with Attention Mechanism

Figure 3

The structural principle of the NMT model based on attention. (a) Conventional NMT. (b) NMT with multichannel encoder.

(a)

(b)