SinoVoice’s application of handwriting recognition software for PC

SinoVoice as a high-tech corporation has acquired many core technologies in the areas of TTS (Text-to-Speech), ASR (Automatic Speech Recognition) and HWR (Handwriting Recognition) and has become one of the few companies possessing multiple HCI (Human Computer Interaction) technologies in China, due to its years of dedication to the research of HCI. SinoVoice has successfully industrialized its technologies in China.

This application manual will introduce how the handwriting recognition technology can be applied on PC.

1. Application area
This software can be applied to all kinds of pc desktop products and office automation products.

2. Functions
1) Handwriting recognition

Handwriting input software can input any recognized character into all kinds of document, including Microsoft Office series software, MSN, QQ, etc. The user does not need to learn a special input method or memorize any Chinese input code. He can write as he thinks. Thus people can transfer their writing habit onto pc. The handwriting input method is more natural and more suitable for Chinese people’s writing habit. The handwriting input software can accurately recognize any written Chinese, English words, numbers, symbols, casually written words and characters written in wrong strokes. It can also provide the user some similar characters, homophones, associated words to select. At the same time, it can transform between simplified Chinese and traditional Chinese characters. The recognition volume has reached national standard of GB18030.

2) Text-to-Speech (TTS)
With SinoVoice’s latest TTS technology, the computer can output the files being input or existed files in audio format, so that the user can proofread the input content in real time just by listening to it. The fatigue of using eyesight can be greatly relieved and the work efficiency can be greatly enhanced.

3) Personalized Signing function
The software enables people to Sign document in their personal handwriting and reserve the original signature. The user can sign electronic files, realize no-paper office work and pass on letters in personal handwriting electrically at any time.

3. Introduction to the products
1) <InputEasy Chinese intelligent input/output system>
InputEasy provides a unified interface to the user, which combines the functions of writing, listening and proofreading all in one and does not require the user to learn to use other software. InputEasy is a system-level program like other coding input methods, such as Wubi input and Pinyin Input method. It has totally changed the traditional way of bounded methods. The user can use all input methods at any time. InputEasy support s GBK21003 and pressure sensing handwriting, which makes the handwriting more lifelike and vivid.

2、<EasyInput Handwriting Input method>
EasyInput has adapted SinoVoice’s latest HWR and TTS technologies. It is developed as input software with concise interface. This software not only supports mouse input, but also supports touchpad input for laptops, which reduces the demand for other accessory hardware.
Comparison of the two software：
A、Similarity：They have similar functions, that they both support handwriting input, TTS and signing function
B、Difference：
ⅰ Product formation：InputEasy is an intelligent input/output system, while EasyInput is handwriting input software, which is based on Microsoft input system.
ⅱ Display mode：After installing InputEasy, there will be an icon on the desktop as a shortcut access. After installing EasyInput, there will be an added input option in the list of ‘language input list’ in the lower right corner of the desktop, which is ‘EasyInput’ handwriting input software. It is showed as the following picture:

<InputEasy Intelligent Input/Output system>

<EasyInput Handwriting Input software>
ⅲ Interface style：InputEasy has two interfaces, edition interface and simple interface. The edition interface is more comprehensive with more powerful edition functions. EasyInput has a concise interface, which is more suitable to edit documents, such as Word, etc. EasyInput is more suitable for touchpad of laptops. The interfaces are showed as follows:

<InputEasy Intelligent Input/Output system>——edition interface

<InputEasy Intelligent Input/Output system>——simple interface

<EasyInput Handwriting input software>——the first line is the handwriting cursor bar, the four lines in the middle show the optional characters, the last line is the state bar.

4. Core Technologies
1) Handwriting recognition technology
Handwriting recognition is the process of transferring orderly trail information, which is formed while writing on a writing device, into Chinese in-code. It actually is a mapping process from coordinate sequence of writing trails to Chinese in-code.
Technical characteristics：
 High recognition rate：99.9% for standard character，99% for casually written character，96.8% for character written in wrong stroke orders
 Wide recognition scope：Chinese (Simplified 6763+ Traditional 5401+ GBK 21003）、numbers, English and other special symbols
 High recognition speed，amounts to 0.001 second /character～0.01 second /character
 Little occupied space：the program takes space of 40-80K
 Flexible adaptation：it can adapt to Windows 2000/XP/Vista、Win CE、Hopen、Palm OS and other pc and embedded operation systems, and many kinds of CPU chips.

2) Text-to-Speech
TTS is also called Text-to-Speech，which is a technology to transfer any combination of text files into audio files by computer and plays the voice by sound card, telephone card and other multimedia apparatus. Simply put it, TTS can make the machine read out the text files.
SinoVoice’s TTS technology adapts the core technology based on ‘adding voice library congressing’ algorithm and ‘tone changing’ algorithm, with the character transforming accuracy of over 99%, syllable clarity over 90%, sentence understandability over 95%. It supports mixed reading of Chinese and English. The size of voice library can be contracted or extended, with the smallest size around 500KB-1.5M.

3) Automatic Speech Recognition
Automatic Speech Recognition is a high technology that can make machine transfer voice signals into correspondent files or commands by let machine recognize and understand the signals.
The accuracy rate amounts to 96% within 200 words. It supports Windows、Win CE、Linux and other desktop, embedded Operation Systems.