Seddiqi, M.Kivrak, H.Kose, H.2024-09-292024-09-292020978-172819136-2https://doi.org/10.1109/ASYU50717.2020.9259859https://hdl.handle.net/20.500.14619/94792020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020 -- 15 October 2020 through 17 October 2020 -- Istanbul -- 165305Using computer-based recognition of hand gestures can enhance interaction between humans and computers. In this paper, the recognition using surface electromyography (sEMG) of stationary hand signs of Turkish Sign Language (TID) is evaluated. We used an armband to obtain sEMG signals and a sliding window-based method to split the data for extracting features. We obtained sEMG and inertial measurement unit (IMU) data from ten subjects (five male and five female) for a selected subset of stationary TID signs. We pre-processed the signals, extracted time-domain and time-frequency-domain features from the sEMG signals. The signs are analyzed for their classification performance with sEMG signals. A random forest classifier for five TID signs is trained on the dataset and achieved 78% leave-one-subject-out cross-validation accuracy for the male subjects. The difference in sEMG signals of males and females is analyzed. The performance of time-domain and time-frequency domain features of sEMG, and the inertial measurement unit (IMU) on the hand gesture recognition are evaluated. © 2020 IEEE.eninfo:eu-repo/semantics/closedAccessGesture RecognitionRandom Forest ClassifierSign Languagesurface EMGConference Object10.1109/ASYU50717.2020.92598592-s2.0-85097931512N/A