IMPLEMENTASI KOMUNIKASI DUA ARAH PADA SERVICE ROBOT MENGGUNAKAN ALGORITMA TRANSFORMERS

DAFFA, SYAUQI ZALFFA and Dwijayanti, Suci (2024) IMPLEMENTASI KOMUNIKASI DUA ARAH PADA SERVICE ROBOT MENGGUNAKAN ALGORITMA TRANSFORMERS. Undergraduate thesis, Sriwijaya University.

	Text RAMA_20201_03041282025045.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
	Text RAMA_20201_03041282025045_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (11MB) | Request a copy
	Text RAMA_20201_03041282025045_0030078404_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB)
	Text RAMA_20201_03041282025045_0030078404_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (692kB) | Request a copy
	Text RAMA_20201_03041282025045_0030078404_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (757kB) | Request a copy
	Text RAMA_20201_03041282025045_0030078404_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_20201_03041282025045_0030078404_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (170kB) | Request a copy
	Text RAMA_20201_03041282025045_0030078404_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (407kB) | Request a copy
	Text RAMA_20201_03041282025045_0030078404_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (4MB) | Request a copy

Abstract

The humanoid robot is a type of robot that can assist in various human tasks as a service robot. This robot must have verbal communication abilities to enable two-way communication systems via speech recognition. Previous research indicates limitations in Indonesian language transcription accuracy due to the scarcity of data samples. This study aims to develop a real-time two-way communication system between humans and robots. The dataset used in this study consists of 21 speakers (15 males and 6 females). The speech-to-text system employs a transformer algorithm with the Whisper model, while text-to-speech utilizes Google Text to Speech (gTTS), a Python library, and a command-line interface (CLI) tool interfaced with Google's text-to-speech API. The best-trained transformer model was achieved using 900 training steps. Simulation results show that this model yields word error rates (WER) and character error rates (CER) of 10% and 4% for male speech samples, and 11% and 8% for female speech samples. Real-time testing under quiet conditions with noise levels of 47.1-59.0 dB shows average WERs of 6%, 12%, and 15% at distances of 10 cm, 30 cm, and 50 cm respectively, with corresponding average CERs of 2%, 4%, and 5%. In machine noise conditions with levels of 82.3-91.3 dB, the average WERs for the same distances are 13%, 16%, and 24%, with CERs of 4%, 5%, and 8%. Meanwhile, in noisy conditions with levels of 72.8-84.4 dB, average WERs are 16%, 20%, and 22%, with average CERs of 6%, 7%, and 8% for distances of 10 cm, 20 cm, and 50 cm. This research demonstrates that speech recognition systems can be implemented for communication between robots and humans, enabling robots to respond appropriately to received voice inputs.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Service robot, Speech recognition, Speech to text, Text to speech
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7800-8360 Electronics > TK7816.P39 Electronics T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7800-8360 Electronics > TK7872.T6K47 Transducers-Design and construction, Detectors, Measuring instruments T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7885-7895 Computer engineering. Computer hardware
Divisions:	03-Faculty of Engineering > 20201-Electrical Engineering (S1)
Depositing User:	Syauqi Zalffa Daffa
Date Deposited:	18 Jul 2024 02:50
Last Modified:	18 Jul 2024 02:50
URI:	http://repository.unsri.ac.id/id/eprint/151505

Actions (login required)

View Item