In this project, we have proposed a mechanism of hand gesture recognition using flex sensors and Arduino UNO. The data acquired from the sensors corresponding to different hand gestures are analyzed with the help of different traditional machine learning algorithms. We also proposed an adversarial learning approach that performs better classification in comparison with these traditional learning models. The results of the proposed approach records a commendable accuracy of 88.32% with a precision of 81.77%, a recall of 84.37% and F1-score of 82.78%. The proposed model yields better results in terms of all these four performance metrics. Index Terms—Hand Gesture Recognition, Flex Sensor, Adversarial Machine Learning, Neural Network.