Tagging Like Humans: Diverse and Distinct Image Annotation

Tagging Like Humans: Diverse and Distinct Image Annotation

Baoyuan  Wu, Weidong  Chen, Peng  Sun, Siwei  Lyu, Bernard  Ghanem, 
"Tagging Like Humans: Diverse and Distinct Image Annotation", 
Conference on Computer Vision and Pattern Recognition (CVPR 2018)
Baoyuan Wu, Weidong Chen, Peng Sun, Siwei Lyu, Bernard Ghanem
image annotation
2018
In this work we propose a new automatic image annotation model, dubbed diverse and distinct image annotation (D2IA). The generative model D2IA is inspired by the ensemble of human annotations, which create semantically relevant, yet distinct and diverse tags. In D2IA, we generate a relevant and distinct tag subset, in which the tags are relevant to the image contents and semantically distinct to each other, using sequential sampling from a determinantal point process (DPP) model. Multiple such tag subsets that cover diverse semantic aspects or diverse semantic levels of the image contents are generated by randomly perturbing the DPP sampling process. We leverage a generative adversarial network (GAN) model to train D2IA. Extensive experiments including quantitative and qualitative comparisons, as well as human subject studies, on two benchmark datasets demonstrate that the proposed model can produce more diverse and distinct tags than the state-of-the-arts.