Speech Recognition for Intelligent System in Service Robots: A Review

Speech recognition and response system technology in service robot research continues to evolve alongside technical advances and the increasing demand for intelligent automation across industries. The development of machine learning, digital signal processing, and artificial neural networks has brought significant advancements to speech recognition systems. While speech recognition has been implemented in various systems, there is limited discussion on categorizing these systems by examining four aspects: speech recognition applications, datasets used, emotional speech recognition, and automatic speech recognition. In this study, a review of these four aspects is conducted, focusing on the algorithms and methods used, along with their advantages and disadvantages, particularly in the context of implementing them in service robots using the Indonesian language. This study involves a literature review of 100 research papers from the last five years. The review concludes that the use of the transformer method combined with other techniques and Indonesian language datasets is still very limited. Therefore, future research will focus on developing and implementing these methods in humanoid service robots.