Gesture recognition system based on CNN-IndRNN and OpenBCI

Surface electromyography (sEMG), as a key technology of non-invasive muscle computer interface, is an important method human-computer interaction. We proposed CNN-IndRNN (Convolutional Neural Network-Independent Recurrent Network) hybrid algorithm to analyse sEMG signals and classify hand gestures. Ninapro’s dataset 10 volunteers was used develop the model, by using only one time-domain feature (root mean square sEMG), average accuracy 87.43% on 18 gestures achieved. The obtains state-of-the-art classification performance with significantly reduced model. In order verify robustness compact real¬time recognition system constructed. based open-source hardware (OpenBCI) custom Python-based software. Results show that 10-subject rock-paper-scissors gesture reaches 99.1%.