Flappy Bird Automation using TensorFlow

Abstract

Playing computer games has made a noteworthy changeover to our age s playing style by making the more significant part of them to stick to it. In this way, tackling those diversion issues is a fascinating subject that requires cautious area explicit component definition and an adaptable amusement hypothesis look calculation. So, in this manner, we picked an amusement flappy feathered creature, which includes exploring a winged animal through a bundle of obstacles (pipes) and influence it to make due for quite a while in a steady progression. In addressing this issue, we employed the use of such RL (reinforcing learning) by defining the correct component and after that through the proper activities at every amusement occasion, contributes the pre-emptive specialist who has gotten from a model of CNN (convolutional neural system) which acts as a decent defining product while removing elements from their previous feed previews. We extend a general framework to learn essential highlights of diversion and deal with the problem as appropriate.