Deep Visual Semantic Embedding with Text Data Augmentation and Word Embedding Initialization

Table 6

Experimental results with word embedding initialization.

Dataset	Feature (s)	Image count				Image retrieval
		R@1	R@5	R@10	Med r	R@1	R@5	R@10	Med r
Flickr8k	Aug	20.9	44.1	58.8	7	14.5	39.0	51.2	10
	Aug + Word2Vec	21.5	49.1	62.3	6	15.1	38.9	53.1	9
Flickr30k	Aug	30.6	57.9	68.5	4	21.4	49.3	61.4	6
	Aug + Word2Vec	33.4	59.2	69.6	3	23.3	49.9	61.7	6