Essentially, this paper goes through the types of gait recognition software that currently exist as well as the reliability of this software. There are a lot of different elements that influence a persons gait (like terrain, shoes, fatigue, etc.), which effects how well the software works. It's a pretty heavy read but it's really interesting if you want to just jump to a certain section and read in depth. This technology is very much out there and in development, and it is definitely being improved upon.
Also, even though this paper is a little bit older, it is still being cited in a lot of more recent research papers and I thought it would give a good enough grasp on the science for our purposes.