How to distinct between far loud voice and near soft voice? How to distinct between far loud voice and near soft voice?
Imagine a voice assistant robot trying to know if a person is commanding with her louder voice from a long distance or with her softer voice from a short distance assuming for both cases amplitudes of the received voices are same. What are the ways to distinguish between these cases?
Any authoritative research work/reference will be helpful.
 A: 
How to distinct between far loud voice and near soft voice?

The simple answer is that sometimes you cannot distinguish between a loud sound produced far away and a soft sound produced close by. This is why headphones work. If you listen to music through good quality stereo headphones, you cannot tell that you are listening to very soft sounds produced close to your ears, rather an orchestra playing at full volume a long distance away.
The whispering galleries in St. Paul's cathedral and other buildings also work because you cannot reliably tell how far away a sound was produced. In this case, soft sounds produced far away are focussed by the shape of the building so that they appear to the listener to have been produced close by.
Of course, stereo recordings, stereo headphones and whispering galleries are specifically designed to fool the listener. If you listen a sound in a "natural" environment then differences in volume/intensity between your two ears, the spread of frequencies and the precise timing of echoes may all give your brain clues as to how far away the sound was actually produced.
A: Maybe you could use the frequency distribution. Low frequencies generally carry better over longer distances discussed here.
