DropRegion training of inception font network for high-performance Chinese font recognition. (May 2018)

• Record Type: Journal Article
• Authors: Huang, Shuangping; Zhong, Zhuoyao; Jin, Lianwen; Zhang, Shuye; Wang, Haobin
• Is Part Of: Pattern recognition. Volume 77(2018:May)
• Page Start: 395
• Page End: 411
• Publication Date: 2018-05
• Subjects: Chinese font recognition -- DropRegion -- Elastic meshing technique -- Inception font network (IFN) -- Text block-based font recognition; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2017.10.018
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A new approach, DropRegion-IFN, is proposed for Chinese font recognition. DropRegion is proposed as a new data augmentation and regularization technique. We design a new deep model named inception font network (IFN). Very high accuracies of 99.78% and 98.98% are achieved for text block and single character font recognition. Abstract: Chinese font recognition (CFR) has gained significant attention in recent years. However, due to the sparsity of labeled font samples and the structural complexity of Chinese characters, CFR is still a challenging task. In this paper, a DropRegion method is proposed to generatea large number of stochastic variant font samples whose local regions are selectively disrupted and an inception font network (IFN) with two additional convolutional neural network (CNN) structure elements, i.e., a cascaded cross-channel parametric pooling (CCCP) and global average pooling, is designed. Because the distribution of strokes in a font image is non-stationary, an elastic meshing technique that adaptively constructs a set of local regions with equalized information is developed. Thus, DropRegion is seamlessly embedded in the IFN, which enables end-to-end training; the proposed DropRegion-IFN can be used for high performance CFR. Experimental results have confirmed the effectiveness of our new approach for CFR.