Fast semantic image retrieval based on random forest

• DOI: 10.1145/2393347.2396344
• Publication Date: 2012-11-14T20:36:17Z
• Authors: Hao Fu; Guoping Qiu
• Keywords: Tree (set theory)

This paper introduces random forest as a computational and data structure paradigm for fusing low-level visual features high-level semantic concepts image retrieval. We use to split the tree nodes labels supervise splitting make images located at same node share similar well similarities. exploit such define neighbor set (SNS) of given union all in leaf that this falls onto. From SNS we further similarity measure (SSM) between two number trees which they within SNS. With SSM, example-based retrieval becomes first finding querying then ranking according SSMs its also show new technique can be adapted keyword-based The inherent efficient leads fast solutions. will present experimental results effectiveness technique.