Scribd
Upload
100 views2 pages

Short Research On Voice Control System Based On Artificial Intelligence Assistant

This paper proposes a low-cost voice control system using a Raspberry Pi, voice recognition module, Google Assistant and IFTTT. Google Assistant recognizes voice commands which are then sent to IFTTT. IFTTT recipes trigger actions, like sending commands to control an API connected to the Raspberry Pi. The system was designed to be affordable and support voice control of various systems using open source tools. Evaluation showed the system could effectively implement voice control using inexpensive hardware and open standards.

Uploaded by

M. Talha Nadeem
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100 views2 pages

Short Research On Voice Control System Based On Artificial Intelligence Assistant

This paper proposes a low-cost voice control system using a Raspberry Pi, voice recognition module, Google Assistant and IFTTT. Google Assistant recognizes voice commands which are then sent to IFTTT. IFTTT recipes trigger actions, like sending commands to control an API connected to the Raspberry Pi. The system was designed to be affordable and support voice control of various systems using open source tools. Evaluation showed the system could effectively implement voice control using inexpensive hardware and open standards.

Uploaded by

M. Talha Nadeem
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Short Research on Voice Control System Based on

Artificial Intelligence Assistant

Tae-Kook Kim
Department of Information and Communications Engineering
Tongmyong University
Busan, Republic of Korea
e-mail: [email protected]

Abstract—This paper proposes a voice control system based the output device. Figure 1 shows the flow of data delivery
on artificial intelligence (AI) assistant. The AI assistant system through Amazon Echo [5]. Users must purchase Amazon Echo
using Google Assistant, a representative service of open API Dot products in order to use the voice recognition system.
artificial intelligence, and the conditional auto-run system,
IFTTT(IF This, Then That) was designed. It cost-effectively
implemented the system using Raspberry Pi, voice recognition
module, and open software. The proposed system is expected to
be applied to various control systems based on voice recognition.

Keywords—Artificial Intelligence (AI); Raspberry Pi; Voice


control; Google Assistant; IFTTT

I. INTRODUCTION
Conventional interfaces involve text or touch inputs. Users
can save costs when using voice recognition (speech
recognition) technology as it provides services quickly through
an efficient interface without requiring many peripheral
Fig. 1. Amazon voice service architecture
devices. Additionally, it increases user convenience with a
simple interface through voice interaction between the user and
device.
Currently, voice recognition technology is rapidly changing III. PROPOSED VOICE CONTROL SYSTEM BASE ON
and continues to be applied to daily-use products. A device that ARTIFICIAL INTELLIGENCE ASSISTANT
supports voice recognition technology is required to use it. This In this paper, a voice control system using Google
paper proposes a voice control system using Raspberry Pi and Assistant, a representative service of open API artificial
open API artificial intelligence (AI) [1, 2]. intelligence, and the conditional auto-run system, IFTTT (IF
This, Then That) was designed. IFTTT is a system that
II. RELATED WORKS connects all apps with devices [7]. It is designed to recognize
voices using Google Assistant and handle voice commands via
Voice recognition system provides a variety of services
the IFTTT platform.
through natural language processing using artificial intelligence
systems equipped through the user's voice commands. The Voice recognition API is a HTTP-based REST API
system is divided into platform-management type and open interface that supports input and output based on the JSON
API type artificial intelligences according to the method of (JavaScript Object Notation) format. To use the voice
artificial intelligence system. Platform-management type recognition API, audio files recorded at a sampling frequency
artificial intelligence is capable of detailed operation, but it is of 16 kHz must be encoded into Base64 and delivered to the
not possible to use services on other platforms as it is open API server via HTTP communication [2].
dependent on a certain platform. For open API type artificial
intelligence, it is possible to use services on other platforms In this paper, Google Assistant was used as the Open API.
[3]. Users use the Google Assistant service to send voice
commands. Voice data sent to Google server are converted to
The representative systems developed in relation to speech text after being recognized; the data are subsequently sent to
recognition technology are Google Home, Amazon Echo, and the conditional auto-run mode IFTTT service. For IFTTT, one
Apple HomePod [4-6]. Once a user sends a voice command, recipe performs one operation. One Recipe created by Trigger
artificial intelligence processes each signal into text, converts and Action sends a PUT request in JSON format to the API
the data processed into speech after returning it, and passes it to
server, which has already been set by itself through Webhooks. effectively implemented the system using Raspberry Pi,
The API server exchanges data via WLAN communication and modules, and open software. This system allows users to
Raspberry Pi. The received PUT request updates the PIN status implement their own system through the advantageous voice
associated with Raspberry Pi. Figure 2 shows the structure of recognition interface and various modules. The proposed
the voice control system proposed in this paper. system is expected to affect the development of the voice
control AI market and related systems.

ACKNOWLEDGMENT
This work was supported by the National Research
Foundation of Korea(NRF) grant funded by the Korea
government(MSIP : Ministry of Science, ICT & Future
Planning) (No. 2017R1C1B2011285).

REFERENCES
[1] Raspberry, [Online] Available: https://www.raspberrypi.org
[2] Electronics and Telecommunications Research Institute (ETRI), Voice
recognition open API, [Online] Available:
http://aiopen.etri.re.kr/guide_recognition.
[3] Choe, Jaeho, and Hoontae Kim. "A Survey Study on the Utilization
Status and User Perception of the VUI of Smartphones." Journal of
Society for e-Business Studies, vol. 21, no. 4, 2017.
Fig. 2. Proposed voice control system structure [4] Google, Google Home. [Online] Available:
https://store.google.com/product/google_home
[5] Amazon, Amazon Echo. [Online] Available:
https://www.amazon.com/Amazon-Echo-And-Alexa-
IV. CONCLUSION Devices/b?ie=UTF8&node=9818047011
[6] Apple, Apple HomePod. [Online] Available:
In this study, the design of a voice control system using https://www.apple.com/homepod
embedded systems and open API AIs is detailed. The proposed [7] IFTTT, IFTTT (if this, then that). [Online] Available: https://ifttt.com
system is a voice command system that operates modules, such
as a relay module, through a user's voice commands. It cost-

You might also like