Zero-shot Handwritten Chinese Character Recognition with hierarchical decomposition embedding

Cao, Zhong; Lu, Jiang; Cui, Sen; Zhang, Changshui

doi:10.1016/j.patcog.2020.107488

Cited by 74 publications

(47 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Analysis on hard cases. We manually select some hard cases (e.g., occlusion, oblique or curved, confusing background, blur, long, and vertical) and analysis the performance of baselines on these cases 9 . Please note that the "Scribble" situation is mainly associated with the handwriting dataset so we do not analyze the "Scribble" case separately.…”

Section: Analysis Of Experimental Resultsmentioning

confidence: 99%

“…According to our observations, only few methods are specifically designed for Chinese text recognition although it has enormous application markets all over the world. In particular, the existing Chinese text methods mainly aim at handwriting text recognition [13,14,19,28,76,79] or handwriting character recognition [9,11,77,80,84,94,95,97] thanks to the detailed specifications formulated by the CASIA-HWDB database [39]. For example, Xiao et al [81] put forward a template-instance loss to distinguish the similar Chinese characters pairs in the feature domain.…”

Section: Existing Text Recognition Methodsmentioning

confidence: 99%

“…However, these recognition methods all work at the character level and they are vulnerable to those characters absent from the training sets. To mitigate the character zero-shot problems, some researchers sought to tackle unseen Chinese characters at the radical level [9,70,74,94]. For example, DenseRAN [74] follows the fashion of image captioning and decodes the radical sequence iteratively.…”

Section: Existing Text Recognition Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Benchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical Study

Chen¹,

Yu²,

Ma³

et al. 2021

Preprint

View full text Add to dashboard Cite

The flourishing blossom of deep learning has witnessed the rapid development of text recognition in recent years. However, the existing text recognition methods are mainly for English texts, whereas ignoring the pivotal role of Chinese texts. As another widely-spoken language, Chinese text recognition in all ways has extensive application markets. Based on our observations, we attribute the scarce attention on Chinese text recognition to the lack of reasonable dataset construction standards, unified evaluation methods, and results of the existing baselines. To fill this gap, we manually collect Chinese text datasets from publicly available competitions, projects, and papers, then divide them into four categories including scene, web, document, and handwriting datasets. Furthermore, we evaluate a series of representative text recognition methods on these datasets with unified evaluation methods to provide experimental results. By analyzing the experimental results, we surprisingly observe that state-of-the-art baselines for recognizing English texts cannot perform well on Chinese scenarios. We consider that there still remain numerous challenges under exploration due to the characteristics of Chinese texts, which are quite different from English texts. The code and datasets are made publicly available at https://github.com/ FudanVI/benchmarking-chinese-text-recognition. Figure 1. Three reasons for the scarce attention of Chinese text recognition. (a) People may use different ways to crop text regions, which leads to unfair comparison. (b) It is necessary to specify the equivalence between lowercase and uppercase, half-width and full-width, simplified and traditional characters. (c) The existing methods are mainly evaluated with English datasets rather than Chinese datasets.

show abstract

Section: Analysis Of Experimental Resultsmentioning

confidence: 99%

Section: Existing Text Recognition Methodsmentioning

confidence: 99%

Section: Existing Text Recognition Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Benchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical Study

Chen¹,

Yu²,

Ma³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…It depends on the functions to be removed. The removed attributes must be able to classify the character habit [18,19]. In this application, the research work have provided a great advantage by using a horizontal cover there is also a maximum limit for functions.…”

Section: Methodsmentioning

confidence: 99%

Enhanced Handwritten Document Recognition Using Confusion Matrix Analysis

Umadevi¹,

Murugan²

2021

Advances in Parallel Computing Technologies and Applications

View full text Add to dashboard Cite

The handwritten Multilanguage phase is the preprocessing phase that improves the image quality for better identification in the system. The main goals of preprocessing are diodes, noise suppression and line cancellation. After word processing, various attribute extraction techniques are used to process attribute properties for the identification process. Smoothing plays an important role in character recognition. The partitioning process in the word distribution strategy can be divided into global and local texts. The writer does not use this header line to write the text which creates a problem for skew correction, classification and recognition. The dataset used are HWSC and TST1. The tensor flow method is used to estimate the consistency of confusion matrix for the enhancement of the text recognition .The accuracy of the proposed method is 98%.

show abstract

“…e system successfully read the text from TV screen and used in a functional verification system. Considering the Chinese character recognition, in the past, several researchers focused their attention on recognition of printed Chinese characters [22,23], handwritten Chinese characters [24,25], characters in the vehicle's license plates [26,27], and recognition of Chinese characters written in calligraphic styles [28].…”

Section: Related Workmentioning

confidence: 99%

MyOcrTool: Visualization System for Generating Associative Images of Chinese Characters in Smart Devices

Rai

2021

Complexity

View full text Add to dashboard Cite

Majority of Chinese characters are pictographic characters with strong associative ability and when a character appears for Chinese readers, they usually associate with the objects, or actions related to the character immediately. Having this background, we propose a system to visualize the simplified Chinese characters, so that developing any skills of either reading or writing Chinese characters is not necessary. Considering the extensive use and application of mobile devices, automatic identification of Chinese characters and display of associative images are made possible in smart devices to facilitate quick overview of a Chinese text. This work is of practical significance considering the research and development of real-time Chinese text recognition, display of associative images and for such users who would like to visualize the text with only images. The proposed Chinese character recognition system and visualization tool is named as MyOcrTool and developed for Android platform. The application recognizes the Chinese characters through OCR engine, and uses the internal voice playback interface to realize the audio functions and display the visual images of Chinese characters in real-time.

show abstract

Zero-shot Handwritten Chinese Character Recognition with hierarchical decomposition embedding

Cited by 74 publications

References 15 publications

Benchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical Study

Benchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical Study

Enhanced Handwritten Document Recognition Using Confusion Matrix Analysis

MyOcrTool: Visualization System for Generating Associative Images of Chinese Characters in Smart Devices

Contact Info

Product

Resources

About