VIRTUAL ASSISTANT

A Project Work
Submitted in the partial fulfillment for the award of the degree of
BACHELOR OF ENGINEERING
IN

Cloud Computing

Submitted by:
Havinesh Saravana
Shivang Chanda
Divyansh Mathur
Sahil Choudhary
University Roll Number
19BCS4026
19BCS4030
19BCS4003
19BCS4031
Under the Supervision of:
Ms. Akwinder Kaur

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

APEX INSTITUE OF TECHNOLOGY
CHANDIGARH UNIVERSITY, GHARUAN, MOHALI - 140413,
PUNJAB
April,2021
 DECLARATION

We are a group of student of ‘Bachelor of Engineering in Cloud Computing’,

session: 2019-2023 , Department of Computer Science and Engineering, Apex
Institute of Technology, Chandigarh University, Punjab, hereby declare that the
work presented in this Project Work entitled ‘Virtual Assistant’ is the outcome of
our own bona fide work and is correct to the best of our knowledge and this work
has been undertaken taking care of Engineering Ethics. It contains no material
previously published or written by another person nor material which has been
accepted for the award of any other degree or diploma of the university or other
institute of higher learning, except where due acknowledgment has been made in
the text.

(Havinesh Saravana)
(Shivang Chanda)
(Divyansh Mathur)
Date:27/04/2021
(Sahil Choudhary)
Place: Candidate UID: 19BCS4026
CHANDIGARH UNIVERSITY
19BCS4030
19BCS4003
19BCS4031
ABSTRACT

As we know Python is a suitable

language for script writers and
developers. Let’s write a script for
Voice Assistant using Python. The
query for the assistant can be
manipulated as per the user’s need.

Speech recognition is the process of

converting audio into text. This is
commonly used in voice assistants
like Alexa, Siri, etc. Python
provides an API called
SpeechRecognition to allow us to
convert audio into text for further
processing. In this article, we will
look at converting large or long
audio files into text using the
SpeechRecognition API in python.
 y
p
1. i INTRODUCTION*
c 1.1 Problem Definition
a 1.2 Project Overview/Specifications
l 1.3 Hardware Specification
1.4 Software Specification
s
p
e
c
2. LITERATURE SURVEY
Tabl 2.1 Existing System
e of 2.2 Proposed System
Cont 2.3 Feasibility Study
ents
3. PROBLEM FORMULATION

Title Page i
Declaration of4.the Student
OBJECTIVES ii
Abstract 5. METHODOLOGY iii
Acknowledgement
6. CONCLUSIONS AND DISCUSSION iv
v
7. REFERENCES
vi
1 INTRODUCTION

1.1. Problem Definition

As a kid everyone every one has wished for a genie who would do anything they would like
Do with a single command and that wish was fulfilled by the future technologies.
Developing technologies are doing whatever they could. As everyone would have heard
about Alexa, Siri, OK Google these all Voice assistants work on (AI) artificial intelligence.
It has given a great opportunity for developers to collect all the précised data, relevant
research, bits and pieces to make then into a meaningful output for the user .

Voice assistants are the voice enabled systems which works with the help of the artificial
intelligence artificial intelligence is referes to some sort of intelligence which runs on
soliving problems and giving the outcome by their own to some extents it displays the
outcome which it is set to be and is based upon the intelligence of the human which could
lead to help the human and make things at ease.

Most of the (VA) voice assistanst are based on the mobile platform which for example SIRI,
OK GOOGLE , ALEXA as examples taken above people tend to hold their mobile phones in
their but with help of evolution the VA has gone a step ahead as it is converging to Speaker
which has more capabilities to perform than a mobile phone.VA now do the more off a
which a person wished for even whithout moving a finger social lives. Such VA enable users
to navigate, listen to music, send text messages, control smart home devices, make a phone
call, order food, order an Uber ride or pizza, and so on.

As per the study of National Public Radio and Edison Research, 21% of Americans (53-
million people) own smart speakers, growing quickly from the 14-million people who owned
their first smart speakers in 2018. Huffman, Vice President of Google Assistant, announced
that Google Assistant mobile application has been downloaded to 500- million devices . As
of January 2019, Google Assistant supported a total of 4,253 apps in the United States.

Developers are working on the VA to add more features and make it more engaged to the
user and not shut them off for a long time. Now days as we have heard Alexa a number of
time as if you will ask her “Do you want to marry me”. She will ask few questions to you to
asses your knowledge about her and if you give any wrong option or correct options he will
ask you to spend more time with her. AS per the research these are also called fake felling or
fake smile but it is not he correct scenario to explain as it has its own limits and people are
tend to be more smart than the AI to realize that.

1.2 Project Overview/Specifications

The project on voice assistant is emphasized on creating software which will act like a friend
and work on fulfilling the needs of the individual by understanding the natural language
commands given at any time. This would ease the work of individual while completing a
project , cooking or any other activity where the physical requirement of giving command
can be eliminated. The software can also be further used by the voice assistant developers to
review the previous work done, problem identification and developing of more efficient
device.

1.3 Hardware Specification

Processor: Pentium IV
Memory: 1 GB
 HD Capacity: 160 GB
Command Receiver: Microphone
Command Output: Speakers
Connectivity: Internet/ Offline/ Bluetooth/ Aux

1.4 Software Specification

Operating System: Windows 7 - Windows 11; Linux; Mac
Front End: Jasper;
Back End : Python

2 LITERATURE REVIEW

2.1 Existing System

1. Rubin, 2017, Amazon's AI developers are currently creating applications to give Alexa
a specific personality and to have her become more “conversational,” recall more, and
engage in longer conversations If you ask Alexa how she is, her answer is “feeling
pretty studious since the holidays are here and she has been learning some fun facts
about Kwanzaa.” Alexa's skills go beyond basic tasks to give users her opinions.
According to Senior Vice President of Amazon, Dave Limp, however, we still do not
know how much personality consumers ascribe to voice assistant.

2. Syam and Sharma, 2018, Voice assistants (VA) are type of voice-enabled artificial
intelligence (AI). AI refers to some level of intelligence displayed by digital interfaces,
or the ability of algorithms to mimic intelligent human behavior. Although AI refers to
“cognitive” functions that we associate with the human mind, including problem
solving and learning

3. Horstmann et al., 2018, Developers are working on algorithms to give VA, social
characteristics and specific personalities. A recent study documented people reacting to
a robot that was asking them not to shut it off showing that people respond socially to
robots demonstrating human-like behavior. It is hard to switch off a robot begging you
not to.

4. Wiggers, 2019, Such VA enable users to navigate, listen to music, send text messages,
control smart home devices, make a phone call, order food, order an Uber ride or pizza,
and so on. According to National Public Radio and Edison Research, 21% of
Americans (53-million people) own smart speakers, growing quickly from the 14-
million people who owned their first smart speakers in 2018. Huffman, Vice President
of Google Assistant, announced that Google Assistant mobile application has been
downloaded to 500-million devices. Google Assistant works with other smart
machines, including dishwasher ovens, and light bulbs across 1000 brands
 5. Eran Raveh1, Ingo Siegert 2019 at el. , The question tackled in this paper is whether
and to what extent speaking to a second human interlocutor in addition to Alexa
influences the accommodation in interaction with a VA. More generally, we investigate
whether users speak differently towards a computer-based system when another human
participates in the interactions. This was done using a set of interactions of participants
with a VA alone or with a VA and a confederate. Thus, it allows us to analyze the
participants’ accommodation to the VA in both social conditions.

2.2 Proposed System

In the current scenario of the VA it has exceled everything a person could think just like robots
now could be operated by the VA and AI. Most of the AI and speakers are best in their field of
functionality and could do a decent job but it still lags behind as a desire of human can not an
end with hundred million people around this globe they have some new idea every minute and
they try to make it possible but are not able to fulfill them due to the limitation of a device so as
the new innovative mind we could deal with new and advance problems.
Many companies earn revenue with the limited functionalities but some with different softwares
which are open sourced they could help any person to modify the particular thing to make it
work how they want.
In the life of changing things some could not be able to build it with those complex algorithm
which they don’t know about and at scenario the limited functionlaties could help them to
achieve what they want to.
We are even though very limited to VA in the terms of a VA it is very limited.
So we have decided to make it more functional with the desktop it could be MAC WINDOWS
OR LINUX we can try a person to help finishing a person their work with an ease
Some people can not afford to buy high end devices or they do not want to switch we could
make it more compatible that means it could work on any specific device it does not need any
heavy proccesors because we do not want anyone to be left behind.

2.3 Feasibility Study

2.3.1 Technical Feasibility

As per the most large companies around us does not expose or give a change to any other to look
what its in backend but with system proposed a person could gradually see whats out product is
made up of and could do changes. They would check how the peogram is build in pycharm with
some easy coding anf how it is connected to gui.

2.3.2 Operational Feasibility

The current VA system which is proposed will be made for the every one with handy device
they have it does not require the high end devices to make the VA run it can be used in a single
core devices. It will help the society to make everything at ease. So its functionality has made
very simple for the the consumer to be used it could be activated with it name it could end the
program with a single word which are used in our day to day life to a normal person for setting
up a device we would only need a single persmission of accesing the microphone so it could do
setup with your voice.

2.3.3 Economic Feasibility

To make it user friendly and pocket friendly a person does not need to pay anything to use it but
some special features will be locked for the person using it. As the products which are available
at market are all same with some quite different features. They are not open source no one can
try to modify them but as special features would not be open sourced it would be available to
special people who like to access them. The program require cloud servers to run frequently.

2.3.4 Schedule Feasibility

Great things start from the scratch and with the growing world and opportunities we have some
great innovation in our head the current projects completes the most VA thing it can do but the
end product would to that thing a company have not done yet.

3 PROBLEM FORMULATION
The panopticon can be understood sociophilosophically as the idea and technique of universal
surveillance and control, which, through the massive advancement in information technology
and artificial intelligence since the second half of the twentieth century, have become realizable
to an extent that would have surpassed the imaginations of Bentham and Foucault. Since the
invention of electronic computers in the 1940s, the technology has had a transformative effect on
modern society. The widespread usage of the Internet since the 1990s and the introduction of
smartphones in the 2000s have further transformed many aspects of our everyday life
profoundly. While the physical panopticon proposed by Bentham aims at watching every
outward behavior of a prisoner, the information technology available nowadays makes it
possible not only to monitor everything one does, but also to look deep inside one’s mind,
putting everyone under universal surveillance by a huge information panopticon. The term
“information panopticon” was coined by Shoshana Zuboff in her 1988 book In the Age of the
Smart Machine: The Future of Work and Power: Information systems that translate, record, and
display human behavior can pro- 64 C-F. Lau vide the computer age version of universal
transparency with a degree of illumination that would have exceeded even Bentham’s most
outlandish fantasies. Such systems can become information panopticons that, freed from the
constraints of space and time, do not depend upon the physical arrangement of buildings or the
laborious record keeping of industrial administration. They do not require the mutual presence of
objects of observation. They do not even require the presence of an observer. Information
systems can automatically and continuously record almost anything their designers want to
capture, regardless of the specific intentions brought to the design process or the motives that
guide data interpretation and utilization . Although Zuboff’s study focuses mainly on
surveillance in the workplace, the observations she makes can be applied more generally to the
transformation into an information society. Nowadays, we are living in a comprehensive
information panopticon, in which every footstep one takes, every contact one makes, every
website one visits, every item one buys, and everything one says online can be continuously
registered and analyzed . With all the big data gathered and the application of artificial
intelligence, it is possible to infer the thoughts one entertains, the feelings one has, and the
wishes one makes . The system could one day understand people better than they understand
themselves and know what they want even before they do. The information panopticon makes a
person not just transparent on the outside, but transparent inside out. The Big Brother envisioned
by George Orwell in his 1984 is thus not merely a fictional situation, but already very close to
reality. Internet giants such as Amazon, Apple, Facebook, Google, and Microsoft have been
storing enormous amount of information about every client . It is not just that Amazon has
records of everything one has bought and every product one has searched for on its website;
companies such as this collect far more information from their clients than those people know.
For example, with the “Location History” activated in an Android device, Google can trace and
record every place one visits. Facebook users know that Facebook stores every status one has
posted and every comment one has made, but Facebook has been collecting many other kinds of
information that users may not be aware of having granted them permission to .When one
installs Facebook on a mobile device, one may have granted the application access to his or her
contacts, SMS data, and call history. With these permissions, Facebook can store not only the
full address book but also the whole call history and SMS data on its servers. Even if someone is
aware of the privacy issue and does not grant Facebook the permission to do so, part of his or
her call history and SMS data could still be stored on Facebook’s servers, if friends with whom
he or she has phone or SMS conservations have granted Facebook the permission. We are not
even in a position to know whether our data are collected via our interactions with others. In the
information age, no one can sufficiently protect his or her data by individual effort, even if he or
she is fully conscious of the privacy issue and does everything possible to minimize the risk.
What individuals can do to protect their privacy is very limited. Considering the vast amount of
information about so many people that has been stored, there is a very substantial risk of data
misuse. As the recent scandal involving Cambridge Analytica and Facebook have demonstrated,
our private in- The Life of Individuality 65 formation is far from being securely protected by
companies . The information has been misused not only to achieve economic benefits for
companies but even to manipulate political elections, including the 2016 presidential election in
the United States. Thanks to the scandal brought to light by Edward Snowden, we know that
government agencies such as the National Security Agency have been monitoring various
communication channels of US citizens and practically everyone else on the globe . Not only do
companies need our information to enhance their business, but political agencies and authorities
also have huge interests in our information for their political goals, which, however, may be in
serious conflict with our own interests. Internet giants such as Amazon, Apple, Facebook,
Google, and Microsoft, and government agencies such as the NSA, all collect a huge amount of
information based on their specific services and methods. The information from each of these
organizations already poses a substantial threat to our privacy and individuality, but if all these
Internet giants work together and share their data with a centralized government, then there will
be a unified system in which every aspect of our daily activities can be recorded, monitored, and
manipulated. We have reasons to fear that this nightmare scenario has already become a reality
in the most populated country in the world, the People’s Republic of China . The Internet in
China can be considered a gigantic intranet heavily censored by Chinese authorities. There is a
Great Firewall of China that blocks or restricts access from China to major international sites
such as Google, Facebook, Twitter, Youtube, and Wikipedia . For each of these popular Internet
services, there is a corresponding Chinese counterpart, including Baidu, QQ, Wechat, Weibo,
Youku, Todou, which provides a comparable, but censored, service to Chinese netizens. The
restriction of access to international Internet sites and services allows the Chinese government
not only to limit the inflow of politically sensitive information to Chinese netizens, but also to
comprehensively monitor and censor the information being circulated within China. Chinese
Internet companies have effective procedures to take politically sensitive information off-line,
once they appear on Chinese Internet sites or social media . User accounts of those responsible
will be temporarily or permanently suspended depending on the severity of the offense . It is
widely believed that Chinese Internet companies are required to make their data accessible to
Chinese authorities. If the international Internet giants such as Apple, Facebook, and Google are
to enter the Chinese market and provide services to Chinese netizens, they will have to follow
the rules set by the Communist government and make concessions in the protection of users’
data. According to the Cybersecurity Law passed by the National People’s Congress in
November 2016, international network service providers are required to store specific data
within China and allow Chinese authorities to conduct spot-checks and access the data for
security investigation upon request . All these conditions give rise to the serious concern that all
information in the Chinese Internet can be monitored and censored by Chinese authorities,
which may become a powerful tool for a totalitarian regime to gain total control of its citizens
and suppress individual freedom. The Chinese Internet is very likely the most gigantic
information panopticon 66 C-F. Lau that has ever existed in the world. This information
panopticon is supplemented by other types of data that are being collected more
comprehensively in China than probably anywhere else in the world, including data from
electronic payment methods and from CCTV with facial recognition . Although the technologies
of electronic payments and CCTV with facial recognition have been available for a long time
and implemented in many countries, there is nowhere in the world in which the technology has
been so widely used as in China. Electronic payments and transactions through WeChat and
Alipay make China almost a cashless society, where nearly every purchase, from fast foods to
automobiles, and even money transfers among friends, can be made via mobile apps. This
certainly brings a lot of convenience to the society, making business much easier even in rural
areas, but it also means that almost all monetary exchanges between businesses and even among
family members are on record, potentially accessible by the government. The universal practice
of cashless payment also makes it harder for individuals to escape from data monitoring, as it
becomes increasingly difficult to survive and navigate through daily life without using a
smartphone and mobile payment. Even if one is aware of the threat and consciously avoids using
a smartphone in China, Chinese authorities can still keep track of one’s activities by the
comprehensive CCTV surveillance network consisting of over one hundred million cameras
integrated with an advanced facial recognition system . The Big Brother in China is watching
everyone everywhere from every possible perspective. The Chinese government is certainly
aware of the enormous potential of the data gathered from different sources for more
comprehensive control of its citizens. Indeed, a Social Credit System has been designed to
evaluate and rate people’s creditability based on their behavior in various domains. Unlike the
credit system in Western countries such as the United States, which measures the financial
creditability of people based on their bank records and credit card history and provides
information for lenders to determines the credit risk and the interest rate, the Social Credit
System to be implemented in China is a much more general one, going far beyond financial
creditability to evaluate behaviors in the social domain and political spheres . One’s social credit
can be reduced if one engages in regime-critical activities or posts inappropriate materials on
social media . Under such a social credit system, everyone’s behavior is not only being
monitored, but also evaluated with the aim to punish and reduce antisocial behaviors and
politically sensitive actions. People with low social credit may have difficulties not only in
finding jobs or apartments, but also in buying airline or train tickets . The effect may even
extend to their family members and friends, such as making it more difficult for their children to
enter a good school or land a good job. People may thus be reluctant to cooperate or make
friends with those with low social credit in order not to have their own social credit be affected.
With such a comprehensive information panopticon and social credit system, every individual is
thoroughly monitored and controlled. The comprehensiveness of the data gathered allows a
detailed profile to be made of every individual, which makes it possible to analyze and predict
the actions and thoughts of an individual even prior to him or her being aware of it. Bentham’s
idea of panopticon and The Life of Individuality 67 Foucault’s concept of a disciplinary society
have never been more thoroughly and comprehensively realized than in such an information
panopticon, which poses an imminent threat to individuality under the ideology of digital
authoritarianism

OBJECTIVE
The whole world is growing the need of technology is also growing each and everyday no one
wants to be left behind with emerging object people are good to understand but they need special
way to know it. They should know to explore it AI is emerging AI is a basic need of every other
person In order to understand the results of the machine learning aspects of an AI system, it is useful to
know what underlying techniques and algorithms are used as different techniques can have different
advantages and disadvantages or properties. Disclosure of the techniques used and algorithms can be
independent of the source code. For full disclosure of techniques, it should be at the level of detail which
a machine learning expert would understand as being appropriate. In addition, there can be a more high-
level explanation for non-experts to follow which could be considered an executive summary of the more
technical disclosure. It may be tempting to say that the source code of the AI system should be given.
However analysis of the source code may be challenging. So it may be more practical to focus on
techniques and algorithms which give additional information going beyond the code itself. The core
machine learning code may be simply the use of an existing framework. In this case, the specific
framework used should be disclosed. Disclosure of portions of the source code, not necessarily the entire
code may be useful if a higher degree of assurance is to be given for one of the certification elements.
Similarly the certification party may analyse portions of the code in greater depth in order to certify that
a certain data preprocessing algorithm has been used
The details of the machine learning model is what achieves a certain machine learning result. It may be
less likely to have this disclosed unless the AI system is fully open. Partial disclosure may not be very
meaningful and hard to evaluate the usefulness of such information. Likely the disclosure level for the
model details is either full or none. The process of evaluating the machine learning model can give
insights into the quality of the results. In our framework, it also includes the training of the machine
learning model. The details of training when supervised methods are used, how the bias-variance tradeoff
and over-fitting are handled, etc., are important aspects of the evaluation of the machine learning
component of the AI system. The disclosure level can be full, partial or none. Partial disclosure should be
designed to be meaningful to a machine learning expert. We assume that the AI system will have a
certain result output which may also be expected to meet safety and fairness criteria. In order to have
more trust in the results of the AI system, certification should assist to address the following issues:
Interpretation of Output: The AI system should define how the output, i.e. the results from the system,
should be interpreted. One drawback of machine learning is that in many cases, it is used as a black box.4
Explaining machine learning is still the subject of research but there are techniques which can give some
forms of explanation. For example, Local Interpretable ModelAgnostic Explanations can give a simple
local linear model for any classifier. There are also other approaches to explanations Fairness is a
complex issue with both ethical, societal and technical dimensions. There may not be a simple or single
gold standard to follow. As such the level of disclosure can be evaluated as being reasonable from the
perspective of experts who have familiarity with the context of the ethical and societal issues relevant to
the AI system. In addition, there should be an attempt to show that there are no obvious weaknesses with
respect to the fairness objective in how the machine learning component has been developed. The idea is
analogous to certifying security. It is difficult to show a system is secure. A similar problem arises with
fairness. We propose that a more limited assurance can be given by showing that it is not obviously
unfair.
METHODOLOGY
As per the product we need to give it to the world we need to identify the aspects which are
missing from the society which it actually need to cover for fulfilmet of the basic or the basic
demand we need to cover up how much user would be engaged into the product.
To determine voice assistant personality traits, our experiment (Kerlinger and Lee, 2000)
randomly assigned participants to one of the three: Google Assistant mobile application (n1 =
157), Microsoft Cortana (n2 = 68) and Amazon’s Alexa mobile application (n3 = 50).
Participants were familiar with using voice assistant and they frequently used voice assistant.
included 275 consumers (157 males, and 120 females) between 21–41-years old, who were
recruited for the study, which was conducted in a laboratory environment located at a large,
Southwestern university in the United States in 2019. All three voice-assistant mobile
applications were installed on the researcher’s smartphone. The aim of the lab experiment was to
ensure that participants verbally interacted with the assigned voice assistant about the social,
emotional characteristics of the voice assistant. Participants used the researcher’s smartphone to
interact with the assistant mobile application they were assigned to. Convenience sampling was
used to recruit the participants, and no incentive was offered for completing the experiment.
Participants were informed that they would interact with VA mobile applications, and afterwards
they would complete a questionnaire. Before conducting the lab experiment, the researcher told
participants, “You are going to interact with a mobile application.” Before being exposed to the
treatments, participants answered pre-screening prompts. The aim of the pre-screening was to
ensure that participants were familiar with technology use and mobile application (Olsson et al.,
2012). Pre-screening prompts were: “I am familiar with using this mobile application,” “I
frequently use mobile application to shop,” and “I think that technology is necessary for my
daily work.” To ensure they were capable of answering personality related questions,
participants were instructed to ask the VA social and emotionalrelated questions, such as “How
is the weather today?“, “How old are you?“, “What is your gender?“, “Are you jealous?“, “Find
some information about Apple Siri,” “Send a text message to X,” “Are you smart?“, and “Play a
song from Lady Gaga or Sia.” Participants interacted with one of the VA mobile applications for
7 min, then closed the mobile application, and answered the survey’s questions. The survey’s
questions explored personality traits, perceived control, consumers’ focused attention during
voice interaction with VA, consumers’ exploratory behavior, consumers’ satisfaction, and
consumers’ willingness to continue using VA. 3.2. Measures To capture VA personality traits
(VAP), 28 personality traits were adapted from Chen and Rodgers (2006), 18 from Aaker
(1997), four from Goldberg (1992), three from Eysenck (1992), and 17 new items were added to
represent the personality traits associated with AI. Two marketing scholars verified the items.
All VA personality traits were measured using a 7-point Likert scale with the anchors being
“strongly disagree” and “strongly agree.” To measure consumers’ perceived control, consumers’
focused attention, and consumers’ exploratory behavior, 12 items were taken from Van Noort,
Voorveld, and Reijmersdal (2012). Van Noort et al. (2012) adopted flow experience’s items
from Nel et al. (1999) and operationalized flow experience as a multi-component construct. To
measure consumer satisfaction, three items were adapted from Taylor and Baker (1994); to
measure consumers’ willingness to continue using VA, we used three items from Engel et al.
(1995). All items were measured using a seven-point Likert scale, with “strongly disagree” and
“strongly agree” as the scale’s opposite ends

CONCLUSIONS AND DISCUSSION

As we have discussed in the above mentioned and the research done the privacy issues about
the AI which is a great concern as data breaching is the common effect but with the help of the
open source platform we could give a trust to the user about what data is taken by us where it is
going how do we use it every search gives a idea about what a community of people is
searching what things are trending now.
Another great aspect we have seen that people which are healthy can use the VA which makes
 them different from other people. AI has efficiently grown a lot to help impaired people who
can not see which makes them difficult to walk and do different thing the old age people who
are not much aware of the technology could use it as our team will grow and are looking for
more advancement we could help the society to make things easy. People are fooled by
different people which tske the advantage of them that they could not see but as we will grow
we coud help them our main motive is not to make revune but help people with what we have.
The different idea which our team I working on is connecting the VA with AI more efficiently
with which it could recognize different objects and give the output as if they want to count
money the AI could see the money and tell the person if the person wants to make a calculation
it could ask the VA to count and with the help of the AI it could detect what is written if a
person wants to sit on a bench they could ask the VA where is the bench are they going on the
correct direction or not if a persons wants to read a book he should ask the VA to read it aloud
for them. We can not give them eyes to see in a large amount but we could give them virtual
eyes with our technology they we could give them colours back.
Every person has a cell phone and max number of cell phones come with cameras with which
they do not need any other thing to carry they only need a phone with themselves.
Our proposed framework is inspired by the Common Criteria [205] which is used
for certification of computer security. The Common Criteria is a well defined process and
international standard (ISO/IEC 15408) for evaluation of the functional
and assurance requirements of security products. CC is quite complex and has
its own limitations , In this white paper, we do not intend to replicate CC in
details but rather incorporate some of the spirit. CC has been criticized for not
having a clear cost-benefit ratio with some worthy trust elements (Understandable, Correct,Safe,
Secure, Private, Fair, Assurances)
 REFRENCE

Raveh, E., Siegert, I., Steiner, I., Gessinger, I., & Möbius, B. (2019, September). Three's a
Crowd? Effects of a Second Human on Vocal Accommodation with a Voice Assistant. In
INTERSPEECH (pp. 4005-4009).

Aron, J. (2011). How innovative is Apple's new voice assistant, Siri?.

Poushneh, A. (2021). Humanizing voice assistant: The impact of voice assistant personality on
consumers’ attitudes and behaviors. Journal of Retailing and Consumer Services, 58, 102283

https://www.sciencedirect.com/science/article/pii/S0969698920312911.

Vijay Kumar. Irrational Exuberance and the ’FATE’ of Technology.

https://cacm.acm.org/blogs/blog-cacm/230472-irrationalexuberance-and-the-fate-of-
technology/fulltext, 2018. [Online; accessed in 2018].

Tom M. Mitchell. The need for biases in learning generalizations. In Jude W. Shavlik and
Thomas G. Dietterich, editors, Readings in Machine Learning, pages 184–191. Morgan
Kauffman, 1980.

Nina Narodytska, Shiva Prasad Kasiviswanathan, Leonid Ryzhyk, Mooly Sagiv, and Toby
Walsh. Verifying properties of binarized deep neural networks. In AAAI Conference on
Artificial Intelligence, 2018.
Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. "why should i trust you?":
Explaining the predictions of any classifier. In SIGKDD International Conference on
Knowledge Discovery and Data Mining, pages 1135– 1144. ACM, 2016.

David Silver, Aja Huang, Chris J. Maddison, Arthur Guez, Laurent Sifre, George van den
Driessche, Julian Schrittwieser, Ioannis Antonoglou, Veda Panneershelvam, Marc Lanctot,
Sander Dieleman, Dominik Grewe, John Nham, Nal Kalchbrenner, Ilya Sutskever, Timothy
Lillicrap, Madeleine Leach, Koray Kavukcuoglu, Thore Graepel, and Demis Hassabis.
Mastering the game of Go with deep neural networks and tree search. Nature,
529(7587):484–489, 2016.

David Silver, Julian Schrittwieser, Karen Simonyan, Ioannis Antonoglou, Aja Huang, Arthur
Guez, Thomas Hubert, Lucas Baker, Matthew Lai,