Analysis of Two AI Applications: Teachable Machine and Quick Draw

This assignment analyzes two artificial intelligence (AI) applications: Teachable Machine and Quick Draw. The analysis begins by describing the functionality of each application, explaining their purpose and how they utilize AI and machine learning. The assignment then evaluates the effectiveness of each application based on accuracy, utility, the methods used, ease of use, and cognitive load. Furthermore, the assignment provides recommendations for improvements and suggests other potential applications, such as Pix2Pix. The conclusion summarizes the importance of AI and machine learning, highlighting the key findings from the analysis of these two applications, emphasizing their strengths and areas for development.

Artificial Intelligence and Machine Learning
1

Contents
1. Choose 2 applications and describe their functionality and why chose them.........................3
2. Evaluate the effectiveness in terms of.....................................................................................4
3. Recommend improvements or suggest other applications......................................................5
4. Summary..................................................................................................................................6
REFERENCES................................................................................................................................7
2

1. Choose 2 applications and describe their functionality and why chose them
Teachable machine- it is based on the tool or platform that help for creating as machine learning
model in fast, easy and accessible to everyone. This kind of tool will help for recognise any kind
of computer image, sounds and possess. Afterwards, it supports for exporting model in particular
sites (Ullah, Al-Turjman and Gagliardi, 2020). Teachable machine is applicable as perform AI
experiment for Google build on the deeplearn.js library. It is one of the best advancement of
machine learning which help for identifying the gesture or trigger events whether it is a sound or
gif. it ensure that train every angles otherwise, it is likely find it hard to generalise in the proper
gesture.
Functionality of Teachable machine
It is consider as flexible- use document files and capture the live examples. It is respectful
of the better way to represent work. If even chose to use then entirely on device, without any
type of microphone, webcam. It automatically recognise the object with the help of their
recognised features.
Quick Draw- it is a kind of online game which is mainly developed by Google. it would
increase the challenges for player to draw a perfect picture of any object and idea about the entire
design. This is the best example of neural network AI (artificial intelligence) in order to guess,
what kind of object drawings (Zhang and et.al., 2020). AI help to learn about the concept of
design pattern in which increasing the ability to guess correctly in future.
Functionality of Quick Draw
Quick Draw tool under the hood and also work as take a sequence model in order to draw
design pattern. Basically, it is used as canvas and feed into combination of different recurrent
network and convolution layers. At last, it is representing the class digits and then will be
generated from output layers.
As per analysis, it has been chosen these two applications because it is mainly used in the
most of cases. These applications are helping to provide the better facilities and services.
3

2. Evaluate the effectiveness in terms of
Teachable Machine
 Accuracy- teachable machine allows to quickly and interactively identified the vision or
system and then recognize the object simply by offering better camera or pressing one of
button. The accuracy of teachable machine is approximately 80%.
 Utility- this can be utilized the tool or resource in which introduce the combination of AI
and machine learning (Terrier and Martin, 2020).
 Method used- teachable machine is basically used the transfer learning method in which
refine all available data or information.
 Ease of use- This kind of teachable machine tool is becoming user –friendly to provide
better instruction and started in proper manner.
 Cognitive load- in teachable machine tool or platform, cognitive tool can be categorized
in Ml.js and P5.js.
Quick Draw
 Accuracy- the Average human reaction time around 0.2 to 0.24 second. In order to
maintain 90% accuracy within Quick draw tools, which means that drawn, aimed and
fired within 0.06 second.
 Utility- The utility of Quick draw is provide the blade access in different positions.
Sometimes, this kind of dual use utility has 75% longer. In order to maintain the better
utility.
 Method used- As bring the rope up to quick draw, combined line method which help to
explore subset of quite large files. It help for accessing the data sets using other technique
or method.
 Ease of use- Generally, it is used as 2-dimensional library in which associated with
Application programming interface. It became easier for individual people to use Quick
draw tool by opening windows, perform the different functions in proper manner.
 Cognitive load- A quick way to count or shown that there are actually 6 different things.
4

3. Recommend improvements or suggest other applications.
Within Quick draw tool, it is important to understand the recognition accuracy and
usefulness in drawing pattern. Therefore, it has been identified both positive as well as negative
aspect, sought feature suggestions (Alimadadi and et.al., 2020). Even if asked about the Quick
draw in which able to recognize to the particular sketched diagram correctly in most of times.
Quick draw’s is consider the simple interface. There are large number of remarked or interface
enabled very fast drawing.
Recommendations-
Initially, it has been developed the tensor-flow based on the Quickdraw application by
using 15 different subjects, instead of consider whole database images. In context of
improvement, it is important for including class, which can easily accessible for any platform or
smartphone devices. Moreover, it would suggest to include some advance features which help
for improving the overall performance of system in proper manner.
Machine learning system are basically increasingly influence many aspect of daily
activities. That’s why, it has been required to create innovative item in which attract people such
as QuickDraw. This tool is accessible by everyone. At certain level, many individuals have faced
the issue due to their functionality in which reflect on the impression of users (Terrier and
Martin, 2020). In this way, it has been suggested to the setup fix time and design pattern within
specific intervals. It was created as fun game which provide the brief idea to represent data
within different formats.
Furthermore, it has been suggested other application such as Pix2Pix, helping to create
model of images. For Example- demo talks to backend sever in which running the flow model. It
is combination of generative adversarial network that help for implementing a proper sketch or
design and become similar to another objects. Sometimes, this kind of application is used in
place of Quickdraw.
5

4. Summary
In above analysis, it has been identified the importance of artificial intelligence, machine
learning and also using into different applications. Analysis the discussion on the basis of two
different applications such as teachable machine, QuickDraw. Identifying the overall
functionality of these AI or machine learning based application. This can help for recognising the
objects and then draws as quick image. Furthermore, it has been identified the effectiveness of
both applications in context of accuracy, cognitive load, utility, methods and ease of use.
However, it has been identified the performance or efficiency of teachable machine and
QuickDraw.
6

REFERENCES
Book and Journals
Alimadadi, A. and et.al., 2020. Artificial intelligence and machine learning to fight COVID-19.
Terrier, R. and Martin, N., 2020, November. A Machine Learning Tool to Match 2D Drawings
and 3D Objects’ Category for Populating Mockups in VR. In International Conference
on Virtual Reality and Augmented Reality (pp. 240-246). Springer, Cham.
Ullah, Z., Al-Turjman, F. and Gagliardi, R., 2020. Applications of artificial intelligence and
machine learning in smart cities. Computer Communications. 154. pp.313-323.
Zhang, H. and et.al., 2020, April. Building Computational Thinking Through Teachable
Machine. In Society for Information Technology & Teacher Education International
Conference (pp. 127-131). Association for the Advancement of Computing in Education
(AACE).
7