RDMMFET: Representation of Dense Multimodality Fusion Encoder Based on Transformer

Table 1

Statistics of data sets used for pretraining.

Image (K)	Questions
	MS COCO (K) [33]	VG (M) [34]	VQA v2.0 (K)	GQA (M) [35]	VG-QA (M) [36]	All (M)
180	617	5.39	658	1.07	1.44	9.18