AI for Youth

No-code AI for Computer

Vision
Legal Disclaimers
The Intel® Digital Readiness Programs and Intel® AI for
Youth program are developed by Intel Corporation.

© Intel Corporation. Intel, the Intel logo, and other Intel

marks are trademarks of Intel Corporation or its
subsidiaries. Other names and brands may be claimed
as the property of others. All rights reserved. Program
dates and lesson plans are subject to change.

Intel technologies may require enabled hardware,

software, or service activation.

No product or component can be absolutely secure.

Results have been estimated or simulated.

Intel does not control or audit third-party data. You

should consult other sources to evaluate accuracy.

Your costs and results may vary.

 Hello

The objective of the program is to create a solution to a real

problem, through Computer Vision.

From this moment you will explore techniques and methods related
to Computer Vision.

Let’s explore No-Code Computer Vision and some popular tools.

3
 You will be able to answer the
following:
1. How to map the steps of the AI Project Cycle in No-Code
Computer Vision solutions?

2. What is the difference between No-Code and Low-Code Computer

Vision?

3. What are some popular No-Code Computer Vison tools?

4
 Module Sections

1. No Code vs Code AI

2. No-code tools for Computer Vision

3. Lobe No-code tool implementation

4. Teachable Machine No-code tool implementation

1. No Code vs
Code AI (for
expert, not
mandatory)
Let’s compare Code and No-code implementing
AI Project Cycle
using Logistic Regression Model

7
 Let’s Begin with Problem Scoping

Problem Data
Evaluation
Scoping Exploration

Data
Modeling Deploymen
Acquisition
t

8
 No-code vs Code: Problem Statement
Problem Statement:

 To classify the three

species of Iris
Flower, which is a
genus of 260-300
species of flowering
plants with showy
flowers

9
Fisher, R. A. (1988, July 1). UCI Machine Learning Repository: Iris Data Set. UCI. [Link]
 No-code vs Code: Problem Statement
The attributes used in
this dataset :

1. sepal length
2. sepal width
3. petal length
4. petal width

10
Eckroth, J. (2018, December 24). Python Artificial Intelligence Projects for Beginners. O’Reilly Online Learning. [Link]
 No-code vs Code: Problem Statement
These are the three species discussed in this dataset of an Iris
Flower

11
Wikipedia contributors. (2022a, June 10). Iris (plant). Wikipedia. [Link]
Let’s use Orange Data Mining a no-code tool!

12
 Data Acquisition

Problem Data
Evaluation
Scoping Exploration

Data
Modeling Deploymen
Acquisition
t

13
 No-code vs Code: Data Acquisition
Initially, let us import and load the Iris data set with code and No-
code tools

Loading Data: Code Loading Data: No-Code

Approach Approach
 No-code vs Code: Data Acquisition
The Iris dataset contains 150 rows and 6 columns

Loading Data: Code Loading Data: No-Code

Approach Approach 15
 Data Exploration

Problem Data
Evaluation
Scoping Exploration

Data Deployment
Acquisition Modeling

16
 No-code vs Code: Data Exploration
We are exploring the data by finding the descriptive statistics of the Iris dataset
such as mean, minimum value, maximum value, standard deviation

Data Exploration: Code Approach Data Exploration: No-Code Approach

17
 Let’s recap Mean
Mean is the average
value of a dataset Example
calculated by adding all
numbers divided by the
set’s number of values.

18
mathsteacher. (n.d.). Mean, Median and Mode. Retrieved June 22, 2022, from [Link]
 Let’s recap Standard Deviation
Std is the
standard
deviation that
measures the
dispersion or
the spread of
the data about
the mean
value.

19
Brainkart. (n.d.). Calculation of Standard Deviation. Retrieved June 22, 2022, from [Link]
 Modeling

Problem Data
Evaluation
Scoping Exploration

Data
Acquisition Modeling Deployment

20
 No-code vs Code: Modeling
For modeling the Iris dataset, we use Logistic Regression, a classification method

Modeling: Code Approach Modeling: No-Code Approach

21
 No-code vs Code: Modeling
For modeling the Iris dataset, we use Logistic Regression, a classification method

Modeling: Code Approach Modeling: No-Code Approach

22
 Evaluation

Problem Data
Evaluation
Scoping Exploration

Data
Modeling Deployment
Acquisition

23
 No-code vs Code: Evaluation
We can evaluate the results by calculating classification metrics such as accuracy

Evaluation : Code Approach Evaluation: No-Code Approach

24
 No-code vs Code: Evaluation
We can evaluate the results by calculating classification metrics such as accuracy

Evaluation : Code Approach Evaluation: No-Code Approach

 Let’s compare Low Code
and No-Code AI with examples

26
 Low-code and No-code Platforms

Low Code No- Code

Orange is a drag and drop tool used for
Python Libraries are used in low-code tool
No-code development

18-Scientific-Python. (n.d.-b). Fabienmaussion. Retrieved August 2, 2022, from [Link]

Knowledge Check!

28
1. The steps in the AI Project Cycle are as follows. State whether it is true or false.

[Link] Exploration

[Link] Acquisition
29
[Link]

[Link] Statement

True OR False

29
1. The steps in the AI Project Cycle are as follows. State whether it is true or false.

[Link] Exploration

[Link] Acquisition

[Link] 30
[Link] Statement

True OR False

30
2. The mean of the following numbers is:
• 6, 10, 12, 19 Choose the answer.

11.75
31 12.75

11.25 12.25

31
2. The mean of the following numbers is:
• 6, 10, 12, 19 Choose the answer.

32
11.75 12.75

11.25 12.25

32
2. No-Code tools
for Computer
Vision
Let’s begin to learn the tools with an introduction
to Computer Vision

34
 Introduction to Computer Vision
 Computer vision, one
of the domains of AI
trains and enables
computers to
understand the visual
world

 Computers use images

and AI models to
accurately identify and
classify the objects

S. (2021i, October 12). What Is Computer Vision : Overview, Applications, Benefits and How to Learn It. [Link]. [Link]
 Why do we use No-code in Computer Vision?
 Reason 1: Orange Data Mining
Tool
 It solves problems without writing a
single line of code, hence makes it
easier to use

 Reason 2:
 It has an easy-to-use interface

 Reason 3:
 The drag-and-drop options allow
the users to visualize the procedure
easily
No-code tools for Computer Vision

These are some examples of No code Tools for CV

37
 No-code tools for CV: Lobe

Lobe works with

image
classification

38
Microsoft. (n.d.). Lobe | Machine Learning Made Easy. Retrieved June 22, 2022, from [Link]
 Applications of Lobe

It allows a set of images

with labels and
automatically finds the
most optimal model to
classify the images

39
Microsoft. (n.d.). Lobe | Machine Learning Made Easy. Retrieved June 22, 2022, from [Link]
 We will be learning more
about Lobe later in this session

40
 No-code tools for CV: Teachable
Machine

Teachable machine is
a web-based tool

41
Google. (n.d.). Teachable Machine. Teachable Machine. Retrieved June 22, 2022, from [Link]
 Applications of Teachable Machine

It allows training of a
model based on different
images, audio, or poses
given as input through a
webcam or pictures

42
Google. (n.d.). Teachable Machine. Teachable Machine. Retrieved June 22, 2022, from [Link]
 We will be learning more about
Teachable Machine in the further sections

43
 No-code tools for CV: Nanonets

 Nanonets build online OCR

(Optical Character
Recognition), which turns
scanned images into text

 It involves Text Recognition,

which identifies the characters
in an image

44
NanoNets. (n.d.). NanoNets - Wiki. Golden. Retrieved June 22, 2022, from [Link]
Knowledge Check!

45
Which is the No-code tool for Computer Vision that has a text
recognition application?

OR
Lobe Nanonets

46
Which is the No-code tool for Computer Vision that has a text
recognition application?

Lobe OR Nanonets

47
3. Lobe No-code
tool
implementation
Let’s understand one No-code AI tool:
Lobe

49
 Introduction to Lobe
• [Link] is an Automl tool,
which means that it is a no-
code AI tool

• It works with image

classification

• It allows a set of images

with labels and it finds the
most optimal model to
classify the images
 Stepwise working of Lobe : Label Data

Label data

Label the data with

variable names, using
images, cameras, or
datasets

51
 Stepwise working of Lobe : Training
Model
Label data

Training the
model

After the data is

labeled and images
are collected, the
training of data
takes place

52
Stepwise working of Lobe : Deploying Model

Label data

Training the
model

Deploying
the model

Evaluate the results and use your

model to check the classification
percentage
53
 Step 0: Getting started with the
website

Go to the website :
[Link]
/

54
 Step 1: Installing Lobe

Initially, we need
to click on
Download to
download the
software

55
 Step 1: Installing Lobe

After Clicking the Download option it shows

the .exe file and gets downloaded

56
 Step 2: Working of Lobe

After downloading
the Lobe software
with the extension
file (.exe), the
Lobe appears in
this manner

Click on New Project

57
 Step 3: Labelling the Data

After selecting the

new project icon,
name the title of
the project

58
 Let’s get started with a hands-on activity
using the Lobe AI Tool for finger counting

59
Step 3: Labelling the Data
On clicking the Import icon,
the formats of image inputs
are displayed

60
 Step 3: Labelling the Data

Select the
Camera Icon
for taking an
image of the
data

61
 Step 3: Labelling the Data
Label 1: three fingers images

Rename the Label to the

name of the image/object

62
 Step 3: Labelling the Data
Label 1: three fingers images The number of images is observed on
the screen after recording i.e., 10
images of 3 fingers

63
 Step 3: Labelling the Data
Label 2: five fingers images

Rename the Label to the

name of the object 2

64
 Step 3: Labelling the Data
Label 2: five fingers images

65
 Step 4: Training the Data
The green color train box indicates that
training of the model is completed

Click on the train icon to

train your model

66
Step 4: Training the Data

The predicted
results are
indicated here
67
 Step 5: Deploying the Model

 To test our model we

need to click on the Use
icon

 The object is placed in

front of the webcam

 The two-color image

boxes are visible, which
shows the classification

68
 Step 5: Deploying the Model

The two boxes

indicate the
correctness of the
image tested

69
Step 5: Deploying the Model
Click on the Export icon to
share your model

70
Knowledge Check!

71
1. From the image below, click on the download option, we have a .exe file to
download. True or False?

72

OR
True False

72
1. From the image below, click on the download option, we have a .exe file to
download. True or False?

73

True OR False

73
2. In Lobe, “Use”, is used during labeling the data. State whether it is
true or false.

OR
True False

74
2. In Lobe, “Use”, is used during labeling the data. State whether it is
true or false.

True OR False

75
Activity: Perform a new Lobe Project
_Divide in teams
_Perform a Lobe Project by choosing a scope
_Some examples: Train the model to distinguish
between healthy and ruined leaves; train the
model to distinguish flowers from leaves

76
4. Teachable
Machine No- code
tool
implementation
 Introduction to Teachable Machine
• Teachable Machine is an AI,
Machine Learning, and Deep
Learning tool developed by Google
in 2017

• It runs on top of
[Link] which was also
developed by the same company

• It is a web-based tool that allows

training of a model based on
different images, audio, or poses
given as input through webcam or
pictures

[Link]: It is a JavaScript library for training and deploying machine learning models in
the browser and in [Link]
78
 Introduction to Teachable Machine
What can we classify in a teachable machine?

Image Audio Pose

Any of the three Data forms can be classified in a teachable machine

79
 Introduction to Teachable Machine
What are the features of a teachable machine?

No-code tool Trains the data within the browser

itself

80
 Introduction to Teachable Machine
What are the requirements to get started working on a teachable
machine?

Computer Webcam Headset with mic

81
Stepwise working of Teachable Machine: Gathering
Data

Gathering
Data

 Gather and group your

examples into classes, or
categories, that you want the
computer to learn

 Upload your image files, or

capture them live with a mic
or webcam

82
 Stepwise working of Teachable Machine: Training
Model
Gathering
Data

Training the
Model

Click on- train model, and

[Link] starts
training a neural network
in your browser

83
Stepwise working of Teachable Machine: Deploying
Model

Gathering
Data

Training the
Model

In the output, you can see the model classification

with percentage and export the model with a
Deploying
sharable link
Model

84
Let’s get started with a Teachable Machine

85
 Step 0: Getting started with the
website

Go to the website :
[Link]
[Link]/

86
Step 1: Working of Teachable Machine

Click on the Get

Started button to
try working with
the software

87
Step 1: Working of Teachable Machine

Shows a New
Project to
begin

88
 Step 1: Working of Teachable Machine

To open an already existing

project we can use any of
these options 2. We can open an
1. Drive already existing
2. File project from a file

1. We can open an
already existing project
from Drive

89
Step 1: Working of Teachable Machine

Select any one form

of the project, such
as image, music, or
pose

90
Let’s get started with a hands-on-activity
with the Teachable Machine to classify
objects using an Image Project

91
Step 2: Working with Image Project

Click on the
image project

92
 Step 3: Selecting an Image Project

The embedded image

model is a smaller
This is a standard image
version of the
model format used for
standard image
taking pictures from a
model, allowing you
webcam or uploading
to run it on
them from a system
microcontrollers
 Step 3: Selecting an Image Project

Click on the blue

highlighted
Standard Image
Model

94
Let’s classify using an example of teddy
bear and tiger toys

95
Step 4: Providing Image Inputs to the Model

 Rename Classes
with the example
of objects chosen
to be classified
 e.g. Class1-toy
 Class2-banana
and so on

96
Step 4: Providing Image Inputs to the Model

Rename Classes
with the objects
Class1-teddybear
Class2-tiger

Note: You can add ‘n’

number of objects using
Add a class

97
 Step 4: Providing Image Inputs to the Model

Rename Classes with the objects

Class1-teddybear
Class2-tiger

98
 Step 4: Providing Image Inputs to the Model
Class 1: Teddy bear images
Click on the upload option
to upload a file of image

Click on the webcam

option to take a picture
99
 Step 4: Providing Image Inputs to the Model
Class 1: Teddy bear images
Click on the upload option to
upload a file of the image

Show the image in front of the webcam and

Click on the webcam option to click on the hold to record option to take ‘n’
take a picture number of images
100
 Step 4: Providing Image Inputs to the Model
Class 2: Tiger images

Similarly,
take
images of
a teddy
bear

In this project, the number of images taken, ’n’=26 (i.e., 26 teddybear images, 26 tiger images)

101
 Step 5: Training the Model

After the images are

taken, click on the
box, with Train
Model

102
Step 5: Training the Model

Once the model is trained, it indicates as Model

Trained

103
 Step 6: Deploying the Model

In the preview box, we

can test the model by
clicking Input in on-
mode and showing the
images of either a teddy
bear or tiger to check
the classification
 Step 6: Deploying the Model
Teddy bear image classification Tiger image classification

105
 Step 6: Deploying the Model

Click on the Export Model

option to see the different
file formats to save your
model

106
 Step 6: Deploying the Model
File Description
[Link] It is a JavaScript library for training and deploying machine
learning models in the browser and [Link]

TensorFlow It is a machine learning model. TensorFlow is an open-

source software library for numerical computation using
data flow graphs

TensorFlow Lite It is a mobile library for deploying models on mobile,

microcontrollers and other edge devices

107
 Step 6: Deploying the Model

Sharable link

108
You can now share this link with your friends
and family if they want to classify teddy
bears and tiger!

109
Let’s look at the procedure for Pose Project
using a doll with positions of being steady
and dancing

110
 Step 2: Working with Pose Project

 Rename the
Classes with
the poses you
have chosen

 Note: You can

add ‘n’
number of
poses using
Add a class
 Click on the Pose
Project icon

111
Step 3: Providing poses to the Model
 Rename the classes to
pose 1- steady
pose 2-dance
 Step 3: Providing poses to the Model
Class 1: Steady Pose

Click on the webcam

option to take a picture

113
 Step 3: Providing poses to the Model
Class 2: Dancing Pose

Click on the webcam

option to take a picture

114
Step 4: Training the Model
The model is getting trained

115
 Step 5: Deploying the Model

 In the preview box, we

can test the model by
clicking Input in on-
mode
 Showing the images
of either a teddy bear
or tiger to check the
classification

116
 Step 5: Deploying the Model

Pose1: steady classification Pose2: dance classification

117
 Step 5: Deploying the Model

 Export your
model using a
sharable link

118
You can now share this link with your friends
and family if they want to classify dancing
and standing poses!

119
Knowledge Check!

120
1. The image shown below is an input for which type of project in a
teachable machine.

121

OR
Audio Project Image Project

121
1. The image shown below is an input for which type of project in a
teachable machine.

122

OR
Audio Project Image Project

122
2. In a teachable machine, we can export our model using a sharable
link. True or False?

123

OR
True False

123
2. In a teachable machine, we can export our model using a sharable
link. True or False?

124

OR
True False

124
 Quiz time

Let’s Kahoot!
 Time to take a short and simple,
fun quiz to see how much have
we learnt!

 Open Kahoot! App or join at

[Link]

 Enter Game PIN to start the quiz!

 Let’s go!
 Discussions
1. List the ethical issues that may be faced while using this tool?
2. Can you list at least two advantages of Teachable Machine over
the Lobe AI tool?

126
 Key Takeaways
 Difference between No-code and Code AI tools

 Various No-code AI tools along with their application for Computer

Vision

 Implementation of Lobe tool

 Implementation of Teachable Machine

127
 Reflection
• Do you think these No-code tools are ethical to use?
• Which tool is easier to use the Teachable Machine or Lobe tools?

128
 Bibliography
 P. (2022b, March 3). AI Platforms with Low-Code and No Code. Build Digital Product Faster, Smarter and Scalable.
[Link]
 Walther, C. (2021b, April 5). What Is OCR? AiThority.
[Link]
nce%20(AI)%20to%20use
 Intelligent document processing with AI | Nanonets. (n.d.-c). Nanonets. Retrieved June 22, 2022, from
[Link]
 How optical character recognition works. (n.d.). Konicaminolta. Retrieved August 2, 2022, from
[Link]
 Get Started | [Link]. (n.d.-c). TensorFlow. Retrieved June 22, 2022, from [Link]
 Sharma, S. (2021c, December 16). What is teachable machine | TEK Society. Medium.
[Link] teachable-machine-a4550bfd3af1
 [Link] | Take Online Courses. Earn College Credit. Research Schools, Degrees & Careers. (n.d.-c). [Link].
Retrieved June 22, 2022, from [Link]
 Tips and Tricks you should know while coding your own Machine Learning Model. (2021b, December 14). Medium.
[Link]
1a83e

 Some Thoughts on Deploying Machine Learning Models | DataMiningApps. (2019b, October 2). Datamining.
[Link]
 Flaticon. (2022, June 17). Webcam Icon - 214713. [Link]
 A. (n.d.). Stereo Headset with Microphone - Adesso Inc ::: Your Input Device Specialist ::: Adesso Inc. Retrieved June 22,
2022, from [Link]
129
 Bibliography
 DCSIL. (n.d.-c). TensorFlow vs [Link] | What are the differences? StackShare. Retrieved June 22, 2022, from
[Link]
s%20and%203%20developer%20stacks
.
 TensorFlow Lite | ML for Mobile and Edge Devices. (n.d.-c). TensorFlow. Retrieved June 22, 2022, from
[Link]
 Microsoft. (n.d.). Lobe | Machine Learning Made Easy. Retrieved June 22, 2022, from [Link]
 Google. (n.d.). Teachable Machine. Teachable Machine. Retrieved June 22, 2022, from
[Link]
 Attention Required! | Cloudflare. (n.d.). Pixabay. Retrieved June 22, 2022, from
[Link]
 Mehrotra, R. (2022, March 21). Sound check: In an age of strong visual mediums, audio is making a strong
comeback. Financial Express. [Link]
 Upham, B., & Laube, J., MD. (2020, July 6). All About Yoga: Poses, Types, Benefits, and More. [Link].
[Link]
 Verma, N. (2020, September 29). What is No-Code Development? Orgzit Blog.
[Link]
 Google. (n.d.-b). Teachable Machine. Teachable Machine. Retrieved June 22, 2022, from
[Link]
 Computer World. (n.d.-b). [Link]. Retrieved June 22, 2022, from
[Link]

130