PERANCANGAN SPEECH RECOGNITION PADA HUMANOID ROBOT UNTUK KOMUNIKASI DUA ARAH DAN VOICE BIOMETRIC BERBASIS RASPBERRY PI 3 MODEL B
Daftar Isi:
- Humanoid robots have become the main attraction in the field of robotics research which has artificial intelligence and can interact with humans, for example having the ability to communicate in two directions for robot service. The robot designed in this study is a humanoid robot which has a combination of speech recognition and voice biometric. Apart from being able to function as a two-way communication robot, the design of this robot can also be used as a voice biometric-based security robot. The core device of this robot is the Raspberry Pi 3 model B and two Arduino UNO R3 as slaves. The main ability of the robot is designed to be able to make movements according to limited orders through speech recognition using a servo motor, which is a manifestation of the interaction between humans and robots. Whereas voice biometrics uses Mel-Frequency Cepstrum Coefficient (MFCC) combined with Delta and Delta-Delta functions for feature extraction processes, while artificial intelligence uses the gaussian mixture model (GMM) as a machine learning algorithm, used for grouping processes (clustering) speakers. The results of the research on robots that have been designed using a 50% comparison of training data and test data reached an accuracy rate of 99.4% and achieved 99.0% accuracy using a ratio of 80% training data and 20% test data.