Department of Multidisciplinary Engineering in Course Project Paper

Vishwakarma Institute
Puja Chavan, Devansh of Technology,
Lathiya, Pune,
Sejal Kurade, 411037,
Pranali Maharashtra,
Likhitkar, India
Tejas Lokhande
VIRTUAL VOICE ASSISTANT IN JAVA
 struggle with traditional user interfaces, such as elderly or those
Abstract – The development of voice assistants has been ongoing who lack technical skills.
for several years, with significant progress made in recent times.
Voice assistants can be implemented in various programming II.LITERATURE SURVEY
languages, including Java, Python, JavaScript, and C++. But Ankit Pandey et al., “Smart Voice Based Virtual Personal
Java’s strong support for object-oriented programming, Assistants with Artificial Intelligence” proposed an intelligent
portability, memory management, large developer community, voice assistant that is able to organize and manage information,
and security make it a great choice for developing complex and including emails, calendar events, files and documents, and
scalable voice assistant applications. This paper proposes a creating to-do lists. The major aim was to create a "Smart voice
system for developing a voice assistant using the Sphinx and assistant in Python" which helps people to control devices with
FreeTTS libraries in Java. The system is designed to have a their voice (speech), extract files and perform various other tasks
simple architecture, with the Sphinx library used for speech on their desktops. This Project is designed using open-source
recognition and the FreeTTS library for speech synthesis. This software module with visual studio code community backing, so
voice assistant can perform various tasks including opening and support available for future updates. The modular approach used in
closing chrome, playing youtube video, booking a cab etc. this project makes it easy to integrate additional modules and
(Keywords: Sphinx API, Speech recognition, FreeTTS, text-to- functions without interfering with existing system functionality [2].
speech, java).
Nishank M. et al., “Voice Assistant for Visually Impaired People”
I. INTRODUCTION They proposed a system to create a special customizable app for the
visually impaired in which commands are accepted in the form of
Voice assistants have become an increasingly popular tool for
text or speech from the user and the tasks are performed
interacting with technology, enabling users to perform a wide range
accordingly. This app acts as a voice assistant. The application is
of tasks with just their voice. Virtual assistants, also known as AI
created by using JAVA programming language on the open
assistants, are a prominent example of this transformation. These
Android Studio platform. Each and every action that is performed
applications are designed to understand voice commands from
by this assistant is spoke. This special app having various features
users and perform a wide range of tasks by processing them
like messaging, call log feature, notes making feature, web
according to algorithms. Well-known virtual assistants like Alexa,
browsing feature will allow visually impaired users to do their
Siri, Google Assistant, Cortana, and more have already made their
basic daily tasks very easily without any other help [3].
presence felt in the market, revolutionizing the way we interact
with technology [1].
S. J. du Preez et al., “AN INTELLIGENT WEB-BASED VOICE
CHAT BOT” discusses the development of an intelligent voice
This research paper focuses on the development of a desktop voice
recognition chat bot. The paper presents a technology demonstrator
assistant system using Java, specifically using the Sphinx and
to verify a proposed framework required to support such a bot (a
FreeTTS libraries. We are using sphinx API for speech recognition
web service). The integration of technologies can also be used in
provided by Sphinx library. The API provides a set of speech
other applications, particularly when dealing with a need for simple
recognition system such as acoustic modeling, language modeling,
input control accessibility and/ or limited processing capabilities.
and decoding. Also, we have used FreeTTS open-source java-based
The voice input from the user and voice output from this system
software package for converting text to speech. It provides a
gives a simpler experience to the user [4].
simple, easy-to-use interface for generating synthesized speech
from text, using a variety of voices and languages. Along with this
Faruk Lawal Ibrahim et al., “A Systematic Review of Voice
we have used sphinx knowledge base generator which builds a
Assistant Usability” discusses how Voice assistants (VA) are an
consistent set of lexical and language modeling files for Sphinx
emerging technology that have become an essential tool of the
decoders.
twenty-first century. The VA ease of access and use has resulted in
These are some tasks that can be performed by this voice assistant.
high usability curiosity in voice assistants. It states that despite the
1.Web surfing
high acceptance rate on the use of VA, not many studies have been
2.Playing youtube videos
carried out on voice assistants’ usability. Researchers reviewed
3.Booking a cab
studies that used voice assistants for various tasks. This study
4.playing music
highlights the usability measures currently used for voice assistants.
5.plan journey
Moreover, this study also highlights the independent variables used
6.opening desktop apps
and their context of use [5].
7.ordering food
Vishal Kumar Dhanraj et al., “Research Paper on Desktop Voice
Since this system works on voice-based commands it enables users
Assistant” They have discussed about Personal Virtual Assistant
to complete tasks and activities hands-free which is especially
for Windows Using Python. This paper defines the working,
convenient for people with disabilities. This voice assistants can be
problems and limitations of voice assistants. The method of
personalized to an individual's preferences. Along with it this voice
creating a voice assistant without using cloud services, which will
assistants can make technology more accessible to people who
allow the expansion of such devices in the future is described in
this paper. Virtual assistants use NLP to match user voice or text

input with executable commands. The Google's online speech
recognition system is used by this system for converting speech exceptional use of URL address which helped us creating
input to text. e. The modular nature of this project makes it easy to such a voice assistant.
understand and more flexible. More features can be added in the
program without disturbing the functionalities [6].

III. METHODOLOGY/EXPERIMENTAL
A. Problem Statement
To develop a desktop voice assistant in Java that can interact with
users through speech. The virtual voice assistant should be able to
understand natural language input from the user and respond
appropriately and able to play music and movies, tell weather
forecast, efficient searching (Google, Wikipedia), Booking Cabs,
railway and bus, suggesting YouTube video for query, location
services etc.

B. Proposed Methodology
The program uses the Sphinx API for speech recognition and
the FreeTTS library for text-to-speech conversion. The code
defines several functions that perform specific tasks based on
the user's input.
There is one voice i.e., (Kevin) which is available in three Fig.1.
frequency voices out of which (Kevin16) is provided to our
voice assistant.
To install the necessary libraries for the voice assistant, we
need to download and configure the Sphinx API and FreeTTS
libraries. These libraries can be added to the Java classpath so
that they can be imported and used in the program.
The program defines several functions that can be triggered by
the user's voice commands. These functions include opening
Chrome, playing music, opening Spotify, opening a
PowerPoint presentation, booking a cab, planning a journey,
and ordering food.
When the program is executed, the voice assistant will start
listening for voice commands. Once a command is received,
the program will execute the corresponding function.
The assistant will continue listening for commands until the
user says "rest," at which point the program will exit. If
suppose voice assistant is unable to hear you it will tell to you
speak that again until your voice is recognized correctly.

C. System architecture
Fig.2.

IV. RESULTS AND DISCUSSIONS

As we are more focused on the problems mention above in Fig.3.
the literature review, we have accordingly tried to develop
such equivalent solution using Java. As of now no voice
assistants are developed in java such that both way
interaction is possible. Also, we have studied various
 available in Python. But when it comes to deploy code to
application or webpage then it becomes easy with Java.
Also, there is no combination of both way communicating
voice assistants are not available due to above reason.
Also, sometimes it seems very difficult when it comes to
make assistant more intelligent with a smaller number of
code and more versatile. So, we have successfully
developed some of depth facilities which voice assistant
can really be helpful to everyone including developers so
that AI and ML can be worked into it and further more
should be developed and launched can be developed in
online and offline mode and we had succeeded in lot of
manner as we have to find the another ways using Java
concept but it is possible to have voice assistants like
which are developed in Python.

REFERENCES
Fig.4. [1] S. Uke et al., "Virtual Voice Assistant in Python
(Friday)," 2022 IEEE 4th International Conference on
Fig.1: Booking cab - when user says to book a cab or
Cybernetics, Cognition and Machine Learning Applications
ride then according to the choice of user it opens that
(ICCCMLA), Goa, India, 2022, pp. 164-167, doi:
interface for you
10.1109/ICCCMLA56841.2022.9989099
Fig.2: Ordering food – when user says to order food [2] Ankit Pandey et al., “Smart Voice Based Virtual
then according to the choice of user it opens that Personal Assistants with Artificial Intelligence” Artificial &
interface for you Computational Intelligence / Published Online: June 2020
VOL1_ISSUE3_18
Fig.3: Train booking- Opens the booking websites by
[3] Nishank M. et al., “Voice Assistant for Visually Impaired
taking choice from user for train or bus.
People” Sep2019, International Research Journal of
Fig.4: Opening Apps – Presentation, Spotify, Chrome Engineering and Technology (IRJET)e-ISSN: 2395-0056
etc. All the desktop application. Volume: 06 Issue: 09
[4] S. J. du Preez, M. Lall and S. Sinha, "An intelligent web-
Also, it can do the all the simple tasks which other voice
based voice chat bot," IEEE EUROCON 2009, St.
assistants can do and very easy further convert into
Petersburg, Russia, 2009, pp. 386-391, doi:
application and websites.
10.1109/EURCON.2009.5167660.
[5] Dutsinma, F.L.I., Pal, D., Funilkul, S. et al. A Systematic
Review of Voice Assistant Usability: An ISO 9241–11
Approach. SN COMPUT. SCI. 3, 267 (2022).
https://doi.org/10.1007/s42979-022-01172-3
5. FUTURE SCOPE [6] Vishal Kumar Dhanraj et al., “Research Paper onDesktop
Today we see there are lots of voice assistant are Voice Assistant.” International Journal of Research in
available. But, as of now nobody uses it in efficient way Engineering and Science (IJRES) ISSN (Online): 2320-
as they have to develop and should customizable 9364, ISSN (Print): 2320-9356 www.ijres.org Volume 10
according to the users and user need. Also, in java such Issue 2 ǁ 2022 ǁ PP. 15-20
voice assistants are not available as there is problem in [7] Research Paper on Desktop Voice Assistant Vishal
speech recognition as the number of task we declare only Kumar Dhanraj (076) Lokeshkriplani (403) Semal Mahajan
those commands are recognized there is no dataset for (427), Feb 2022.
universal recognition. So we can develop such package in [8] AI Based Voice Assistant Using Python 1Deepak
future for it also some methods can be developed to make Shende, 2Ria Umahiya, 3Monika Raghorte, 4Aishwarya
use of AI and ML in it. so that everyone can use this Bhisikar, 5Anup Bhange 1Student, 2Student, 3Student,
technology very conveniently and efficiently. Also, in 4Student, 5Assistant Professor, Feb 2019.
future we can develop such a voice assistant like Jarvis in [9] An Analysis on Voice Assistant Shravankumar Jha, June
the 'iron man' movie which would be able to do any task 2021.
without using Any hardware device. [10] The SPACE THEA Project Martin Spathelf and Oliver
Bendel School of Business FHNW, Bahnhofstrasse 6.
6. CONCLUSION [11] Using Smart Assistants in The Workplace Small
Business Energy Savings| Feb 2022.
Voice assistants in Java nowadays are really not available
[12] Anxo Pérez, Paula Lopez-Otero, Javier Parapar.
as the barrier of it due to certain functionalities which are
“Designing An Open-Source Virtual Assistant”.
[13] Artificial Intelligence-based Voice Assistant by S
Subhash; Prajwal N Srivasta; S Sidhesh; A Ullas; B Santosh.
[14] An Analysis on Voice Assistant Shravankumar Jha,
June 2021.
[15] Voice Assistants and Smart Speakers in Everyday Life
and in Education George TERZOPOULOS, Maya
SATRATZEMI Department of Applied Informatics,
University of Macedonia, Thessaloniki, Greece.
[16] Anxo Pérez, Paula Lopez-Otero, Javier Parapar.
“Designing An Open-Source Virtual Assistant”.