Research on Video Captioning Based on Multifeature Fusion

Table 3

Comparing the experimental results with the representative research work in the field of video captioning.

Models	BLEU4	METEOR	ROUGEL	CIDEr
MPool [14]	0.304	0.237	0.520	0.350
Ruc-uva [13]	0.387	0.269	—	0.459
S2VT [17]	0.314	0.257	0.559	0.352
TA [16]	0.285	0.250	0.533	0.371
SAAT [21]	0.399	0.277	0.612	0.510
M^3-Inv3 [19]	0.381	0.266	—	—
SGN [22]	0.408	0.283	0.608	0.495
PickNet [12]	0.389	0.272	0.595	0.421
Ours	0.443	0.327	0.619	0.521