Parametric and nonparametric context models: A unified approach to scene parsing
作者：
Highlights：
• A new nonparametric scene parsing approach is proposed.
• Parametric and nonparametric contexts models are combined into a unified framework.
• Parametric model considers the co-occurrence of objects as well as their priority.
• Context knowledge is used to improve the first step of nonparametric scene parsing.
• Label transferring and aggregating steps are also improved using context knowledge.
摘要
•A new nonparametric scene parsing approach is proposed.•Parametric and nonparametric contexts models are combined into a unified framework.•Parametric model considers the co-occurrence of objects as well as their priority.•Context knowledge is used to improve the first step of nonparametric scene parsing.•Label transferring and aggregating steps are also improved using context knowledge.
论文关键词：Scene parsing,Parametric,Nonparametric,Context model,Co-occurrence graph,Unified framework
论文评审过程：Received 13 September 2017, Revised 18 June 2018, Accepted 8 July 2018, Available online 17 July 2018, Version of Record 19 July 2018.
论文官网地址：https://doi.org/10.1016/j.patcog.2018.07.013
