#3 14 Aug 2009 11:13:18 AM |
Threads: 0 Posts: 1
Deobrat___ |
Re: How is speech converted to text in computer.?It is highly based upon corelation. You must have noticed that most voice recognition systems require certain degree of training before they can actually be put to use. What they do is that, they record the way you pronounce the phenomes. A phenome is the smallest unit of sound that humans can produce. After building a complete list of phenomes and the way you can pronounce them, it can start listening to your voice and then by splitting up the phenomes, it can start concatenating the phenomes and do a lookup in dictionary to find out the word that matches the series of phenomes. A lot of other heuristics are involved in recognition of exact word. You should look up some resources on Google Serach for detailed information |
| Quote Reply |
|
|