{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/drawing-and-recognizing-chinese-characters","title":"Drawing and Recognizing Chinese Characters with Recurrent Neural Network","arxiv_id":"1606.06539","date":"2016-06-21","proceeding":null,"authors":["Xu-Yao Zhang","Fei Yin","Yan-Ming Zhang","Cheng-Lin Liu","Yoshua Bengio"],"abstract":"Recent deep learning based approaches have achieved great success on\nhandwriting recognition. Chinese characters are among the most widely adopted\nwriting systems in the world. Previous research has mainly focused on\nrecognizing handwritten Chinese characters. However, recognition is only one\naspect for understanding a language, another challenging and interesting task\nis to teach a machine to automatically write (pictographic) Chinese characters.\nIn this paper, we propose a framework by using the recurrent neural network\n(RNN) as both a discriminative model for recognizing Chinese characters and a\ngenerative model for drawing (generating) Chinese characters. To recognize\nChinese characters, previous methods usually adopt the convolutional neural\nnetwork (CNN) models which require transforming the online handwriting\ntrajectory into image-like representations. Instead, our RNN based approach is\nan end-to-end system which directly deals with the sequential structure and\ndoes not require any domain-specific knowledge. With the RNN system (combining\nan LSTM and GRU), state-of-the-art performance can be achieved on the\nICDAR-2013 competition database. Furthermore, under the RNN framework, a\nconditional generative model with character embedding is proposed for\nautomatically drawing recognizable Chinese characters. The generated characters\n(in vector format) are human-readable and also can be recognized by the\ndiscriminative RNN model with high accuracy. Experimental results verify the\neffectiveness of using RNNs as both generative and discriminative models for\nthe tasks of drawing and recognizing Chinese characters.","url_abs":"http://arxiv.org/abs/1606.06539v1","url_pdf":"http://arxiv.org/pdf/1606.06539v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"drawing-and-recognizing-chinese-characters","repo_url":"https://github.com/YifeiY/hanzi_recognition","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"handwriting-recognition","task_name":"Handwriting Recognition"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1606.06539","atlas_url":"https://app.syntology.ai/?focus=1606.06539","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}