Okay, simple answer is that there is very limited software for actually transcribing from speech-to-text (note what philbert had to say in his response in that link). I do know that there is significant work going into development of software to enable s-t-txt for the storage of audio archives from old(er) magnetic tapes for example.
However, using the information in the link I found, developing iListen to recognise multiple voices is nigh-on impossible owing to the huge range of inflexion and modulation in the human voice.
So, it looks at this stage, and without new information from others in the forum, I'd endorse what philbert had to say - manual transcription. You have my sympathies - I spent the best part of a year, in spare time, transcribing the aural memoirs of an elderly community member. I'm glad I did it, despite the trepidation - and it improved my typing skills no end!!!
What??? Did I type 'aural'? Aaaaaagh! Of course, I meant 'oral'.