Efficient multi-modal fusion on supergraph for scalable image annotation
作者：
Highlights：
• We construct a supergraph to structurally combine various types of visual features.
• The main challenge of learning on supergraph is its large computational time.
• To reach scalability, we conduct learning on a small prototype graph in supergraph.
• Prototype graph is a good replacement for sample graph during label propagation.
• We achieve good performance by reconstructing labels of images from prototypes.
摘要
Highlights•We construct a supergraph to structurally combine various types of visual features.•The main challenge of learning on supergraph is its large computational time.•To reach scalability, we conduct learning on a small prototype graph in supergraph.•Prototype graph is a good replacement for sample graph during label propagation.•We achieve good performance by reconstructing labels of images from prototypes.
论文关键词：Image annotation,Manifold learning,Multi-modal representation,Prototype,Supergraph
论文评审过程：Received 1 September 2014, Revised 8 December 2014, Accepted 20 January 2015, Available online 30 January 2015.
论文官网地址：https://doi.org/10.1016/j.patcog.2015.01.015
