Can speech recognition software transcribe MP3 file and creates an output text file? – If yes, how can software deal with voices included without training voice recognition or deal with noisy environment?
Hello,
A transcription software can recognize speech from any audio file, it separate speech streams from sound and music streams, it’s all about Signal Processing, words spoken by human are converted into signals that are converted also into a coding pattern which meaning has been assigned.
You can find about  How Speed Recognition Works article which aims to discuss the software that is capable of defining the decomposition of Speech/Music/Noise and find robust sensors to achieve each component, indexing speech/ music is based on a differentiated modeling and on the use of spectral indices.
