Gated Object-Attribute Matching Network for Detailed Image Caption

Figure 6

A visualized example of the localized objects corresponding to the detected semantic features. (a) GT: a group of sheep are standing together in the grass. Ours: a group of sheep are gazing on the green grass field. (b) GT: a cat is sitting on the bench. Ours: a black cat laying down on a stone bench. (c) GT: a woman with an umbrella is walking down the street. Ours: A woman walks down the street holding a green umbrella.

(a)

(b)

(c)