首页> 外文期刊>ACM Transactions on Graphics >Sketch Classification and Classification-driven Analysis using Fisher Vectors
【24h】

Sketch Classification and Classification-driven Analysis using Fisher Vectors

机译:草图分类和使用Fisher向量的分类驱动分析

获取原文
获取原文并翻译 | 示例
           

摘要

We introduce an approach for sketch classification based on Fisherrnvectors that significantly outperforms existing techniques. For thernTU-Berlin sketch benchmark [Eitz et al. 2012a], our recognitionrnrate is close to human performance on the same task. Motivated byrnthese results, we propose a different benchmark for the evaluationrnof sketch classification algorithms. Our key idea is that the relevantrnaspect when recognizing a sketch is not the intention of the personrnwho made the drawing, but the information that was effectively expressed.rnWe modify the original benchmark to capture this conceptrnmore precisely and, as such, to provide a more adequate tool for thernevaluation of sketch classification techniques. Finally, we performrna classification-driven analysis which is able to recover semantic aspectsrnof the individual sketches, such as the quality of the drawingrnand the importance of each part of the sketch for the recognition.
机译:我们介绍了一种基于Fisherrnvector的草图分类方法,该方法明显优于现有技术。对于rnTU-Berlin素描基准[Eitz等。 2012a],我们在同一任务上的认可度接近人类的表现。基于这些结果,我们为评估草图分类算法提出了不同的基准。我们的主要思想是,识别草图时的相关人眼不是制作图纸的人的意图,而是有效表达的信息。我们修改了原始基准以更精确地捕获此概念,从而提供更充分的信息。重新评估草图分类技术的工具。最后,我们执行分类驱动的分析,该分析能够恢复单个草图的语义方面,例如图形的质量以及草图各部分对于识别的重要性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号