This project can run on ZYNQ board (ARM+FPGA) with a linux system in ARM, (optional) acceleration process implemented with FPGA.
main.cpp: main function to execute qubit discrimination test with C++ feedforward neural network.
to use: g++ -o main main.cpp & ./main

qubit2lay.py: Python qubit discirmination function with 2 layers neural network, generating weights.txt and bias.txt. The structure of neural network in Python and C++ is exactly the same.

trans.py: transform the origin data from .gzip to txt, so as to be read by main.cpp.

fcl.cpp/fcl.h: fully connected layers computation functions implemented on FPGA instead of ARM.

bright.txt/bright1.txt: testing data, only contains bright states data.