Keybook: Unbias object recognition using keywords
作者：
Highlights：
• Current object recognition systems suffer from dataset bias issue.
• Keybook is proposed to undo the dataset bias problem.
• Mutual information is employed to assist the generation of Keybook.
• Empirically, the proposed method shows improvements over conventional solutions.
• Future work include extension to human motion analysis and fine-grain classification.
摘要
•Current object recognition systems suffer from dataset bias issue.•Keybook is proposed to undo the dataset bias problem.•Mutual information is employed to assist the generation of Keybook.•Empirically, the proposed method shows improvements over conventional solutions.•Future work include extension to human motion analysis and fine-grain classification.
论文关键词：Dataset bias,Object recognition,Codebook generation,Bag-of-Words model
论文评审过程：Available online 17 January 2015.
论文官网地址：https://doi.org/10.1016/j.eswa.2015.01.019
