Hello. I have 3 voice samples of a person and 1 sample from an unknown source. I would like to find out if it was recorded by the same person or not. I know there are some commercial voice recognition, voice authentication software on the market, but I would like to find a free and opensource alternative. Can you help me?
I would have imagined that humans are better suited to this task than software.
But there might be an advantage in that a jury might not know just how bad a job software does of this and, incorrectly, assume the machine in infallible.
Same with facial recognition, if there is a human available, they are better at it, at least for the moment. Assuming you aren't comparing very large numbers of samples.
Software will really struggle if the conditions are challenging, (background noise, highly compressed audio like in a cell phone call, etc..)
Thank you. (I think you are right. Probably I should ask some people to listen to the samples and judge. People must be better than machines at this.)