C# – speech recognition from audio file instead of microphone

.netc++delphispeech-recognition

How can I perform speech recognition on speech coming from an audio file (.mp3, wav) instead of the microphone ?

I want to be able to do that from C#.NET and Delphi.