Use k-means clustering to group seed in seed pool to improve seed selection.
It is same as AFL. Before start afl-fuzz, open a group server.
$ python3 group_seed.py [port]
# Another Terminal
$ ./k-means-fuzz -i testcase_dir -o findings_dir -p [port] -- /path/to/program [...params...]