Unfortunately, I don’t have any practical experience with speech or handwriting recognition. However, I would like to get the experience and to use these new features of Windows 7 when communicating with my computer.
It is included free as a part of the Windows Vista and Windows 7 operating systems. The version in Windows 7 is actually unchanged, though some small improvements – such as an expanded dictionary – have been involved.
When a user wants to implement the function Speech recognition, he must do 3 steps:
The function Speech recognition is available in English, French, Spanish, German, Japanese and Chinese (Traditional and Simplified).
Microsoft has been working on handwriting recognition for over 15 years (from Windows 3.0). There were Only 12 languages are available in the Windows Vista. In the Windows 7 more languages are included and for us it is important that Czech language is in the package. The language, in which the operating system is installed, and English language as well can be always usable for handwriting recognition, but when you need to recognize handwriting in another language besides English, you must have Windows 7 Ultimate and download an additional package.
The reason why all languages, used in countries where Windows may be applied, aren’t included in Windows 7 is simple. For each new language it is necessary to collect samples of native handwriting, to analyze collected data and cleanse it. So, development of a new handwriting recognizer starts with a huge data collection effort. Millions words and characters of a written text are collected from tens of thousands of writers. The problems are that some languages have special characters or accents and people in different regions learn to write in different ways. Differences exist even between countries with the same language as between the UK and US, for example. Characters that may look visually very similar to you can actually be quite different to the computer. This is why it is necessary to collect real data about how characters, punctuation marks and other shapes are exactly written.
Before start of data collecting, recognizer developers configure collection tools, prepare documentation and compile language scripts in the labs. Once tools and scripts are ready, the labs are opened and volunteers may donate their handwriting samples. In the course of samples evaluating a gender, age, left handiness and educational background are taken into consideration. A collection session lasts 60-90 minutes. The donated data is then uploaded and stored in a Microsoft database for future use.
Subscribe with BlogLines
Pages of the blog