Modality-Dependent Cross-Modal Retrieval Based on Graph Regularization

Figure 3

Precision-recall curves of all compared methods on three datasets: (a) I2T on the Wikipedia dataset; (b) T2I on the Wikipedia dataset; (c) I2T on the Pascal sentence dataset; (d) T2I on the Pascal sentence dataset; (e) I2T on the INRIA-Websearch dataset; (f) T2I on the INRIA-Websearch dataset.

(a)

(b)

(c)

(d)

(e)

(f)