Curriculum learning of visual attribute clusters for multi-task classification

作者：

Highlights：

• Novel learning scheme that combines the advantages of multi-task and curriculum learning.

• Efficient correlation-based grouping of tasks based on hierarchical clustering.

• Extensive evaluation, analysis of covariates and ablation studies.

摘要

•Novel learning scheme that combines the advantages of multi-task and curriculum learning.•Efficient correlation-based grouping of tasks based on hierarchical clustering.•Extensive evaluation, analysis of covariates and ablation studies.

论文关键词：Curriculum learning,Multi-task classification,Visual attributes

论文评审过程：Received 14 September 2017, Revised 19 January 2018, Accepted 25 February 2018, Available online 5 March 2018, Version of Record 15 March 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.02.028