Latent topics-based relevance feedback for video retrieval

作者：

Highlights：

• The retrieval problem is addressed as a class discovery problem by latent topics.

• A new ranking function is presented to deal with the semantic gap challenge.

• A retrieval framework is defined using the proposed ranking function.

• The results show the effectiveness of the proposed approach.

摘要

Highlights•The retrieval problem is addressed as a class discovery problem by latent topics.•A new ranking function is presented to deal with the semantic gap challenge.•A retrieval framework is defined using the proposed ranking function.•The results show the effectiveness of the proposed approach.

论文关键词：Content-Based Video Retrieval,Relevance feedback,Latent topics,Probabilistic Latent Semantic Analysis (pLSA),Information retrieval

论文评审过程：Received 6 March 2014, Revised 10 July 2015, Accepted 13 September 2015, Available online 28 September 2015, Version of Record 27 November 2015.

论文官网地址：https://doi.org/10.1016/j.patcog.2015.09.007