Research Article

Visual-Text Reference Pretraining Model for Image Captioning

Figure 3

Calculation process of VRLN.