Graph-based bag-of-words for classification

作者：

Highlights：

• A graph-based model to encode local structures of an object into graphs.

• A bag-based framework to encode graphs into vectors.

• Successful validation of the framework in different pattern recognition applications.

• Efficiency and effectiveness gains of the proposed method when compared with baselines.

摘要

•A graph-based model to encode local structures of an object into graphs.•A bag-based framework to encode graphs into vectors.•Successful validation of the framework in different pattern recognition applications.•Efficiency and effectiveness gains of the proposed method when compared with baselines.

论文关键词：Bag models,Graph matching,Graph-based retrieval,Graph-based classification

论文评审过程：Received 6 September 2016, Revised 25 July 2017, Accepted 8 September 2017, Available online 12 September 2017, Version of Record 28 September 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.09.018