SIBI Syllable Recognition System With LSTM

Deaf people are people who are unable to communicate with spoken language. Deaf people communicate with the Indonesian Sign System or commonly known as SIBI. In fact, deaf people often communicate and socialize with normal people, but not all normal people understand and understand SIBI. We as researchers want to provide a way out so that deaf people can convey information to normal people, who do not understand SIBI. The input of this research is SIBI movement which is saved in Video format, then the video is processed with the LSTM algorithm. Video The results of the processing are in the form of classified words, the output of this research is in the form of text. The word data set in this study consisted of 3 words, each word contained 30 data sets. The level of accuracy of this study when recognizing words in SIBI that do not hit the face is 100%.