Effective voice recognition in real-world environments is hard for at least two reasons. First, the voice-recognition system has to be able to recognize words with a very high degree of accuracy. Second, the system has to be able to identify and isolate the speaker in environments where there is a good deal of ambient noise. Microsoft has recently published a technical paper in which they present a system that is capable of solving the first problem. VocalZoom has developed an intriguing approach to solving the second.
Read more on Forbes here.