Stroke classification for sketch segmentation by fine-tuning a developmental VGGNet16

Stroke classification for sketch segmentation by fine-tuning a developmental VGGNet16
复制标题

通过微调开发 VGGNet16 进行草图分割的笔画分类

DOI:
10.1007/s11042-020-08706-y
复制
发表时间:
2020-03
影响因子:
3.6
通讯作者:
Qin Zheng
Qin Zheng
中科院分区:
计算机科学4区
文献类型:
--
作者:
Zhu Xianyi;Yuan Jin;Xiao Yi;Zheng Yan;Qin Zheng

文献摘要

参考文献

相似文献

草图分割和标注面临两个挑战:样本少和特征少。3D数据驱动的方法使用附加标记的3D网格来增加采样。然而,对于没有对应的3D网格的抽象草图,它们是不可行的。而基于特征的手工方法虽然不需要三维网格,但对各种笔画都很敏感。为了应对这些挑战,我们探索了基于卷积神经网络(CNN)的迁移学习,通过微调预训练的CNN来对草图分割的笔画进行分类。我们提出了一种新的信息输入的CNN,使笔划的位置信息清晰。为了改善迁移学习过程中的微调,我们建议将分组过滤层添加到CNN中,使CNN的表示能力增加。实验结果表明,与现有技术相比,在抽象草图数据集上的性能提高了9.7%,在具有相应三维网格的草图数据集上的性能提高了2%。
Sketch segmentation and labeling face two challenges: few samples and few features. 3D data-driven methods use additional labeled 3D meshes to increase samples. However, they are not feasible for the abstract sketches that have no corresponding 3D meshes. And handcrafted feature based methods, although need no 3D meshes, are sensitive to various strokes. To address the challenges, we explore transfer learning based on convolutional neural network (CNN) by fine-tuning a pre-trained CNN to classify strokes for sketch segmentation. We propose a novel informative input for the CNN, making the position information of strokes clear. To improve fine-tuning during transfer learning, we propose to add grouped filter layers to the CNN, making the CNN’s representational capacity incremental. Compared with the state-of-arts, our experimental results achieve 9.7% improvement on the abstract sketch dataset, and 2% improvement on the sketch dataset that has corresponding 3D meshes.
DOI: 10.1007/s11042-017-4987-0
发表时间: 2017-07
影响因子: 3.6
作者:
Liang Wan;Yi Xiao;Ning Dou;Chi-Sing Leung;Yu-Kun Lai
通讯作者: Liang Wan;Yi Xiao;Ning Dou;Chi-Sing Leung;Yu-Kun Lai
DOI: 10.1007/s11042-016-4187-3
发表时间: 2017-12
影响因子: 3.6
作者:
Bo Li;Yijuan Lu;H. Johan;Ribel Fares
通讯作者: Bo Li;Yijuan Lu;H. Johan;Ribel Fares
DOI: 10.1109/iccv.2015.144
发表时间: 2015-05
期刊: 2015 IEEE International Conference on Computer Vision (ICCV)
影响因子: --
作者:
Songfan Yang;Deva Ramanan
通讯作者: Songfan Yang;Deva Ramanan
DOI: 10.1145/3272127.3275051
发表时间: 2018-12
期刊: ACM Transactions on Graphics (TOG)
影响因子: --
作者:
Changjian Li;Hao Pan;Yang Liu;Xin Tong;A. Sheffer;Wenping Wang
通讯作者: Changjian Li;Hao Pan;Yang Liu;Xin Tong;A. Sheffer;Wenping Wang
DOI: 10.1007/s00158-007-0093-7
发表时间: 2007-02
影响因子: 3.9
作者:
T. Lewiński;G. Rozvany
通讯作者: T. Lewiński;G. Rozvany