Chapter -1

Introduction
1.1 Cloaking Technology
A cloaking device is a hypothetical or fictional stealth technology that can cause objects, such as
spaceships or individuals, to be partially or wholly invisible to parts of the electromagnetic (EM)
spectrum. However, over the entire spectrum, a cloaked object scatters more than an uncloaked
object. Fictional cloaking devices have been used as plot devices in various media for many
years. Developments in scientific research show that real-world cloaking devices can obscure
objects from at least one wavelength of EM emissions. Scientists already use artificial materials
called meta-materials to bend light around an object.
“Invisibility” has not remained a thing of science fiction movie or story. Imagine when one will
be able to disappear in the air and no one is able to see him like the Harry Potter invisibility
Cloak. Scientist has discovered the devices which are able to bend the light around three
dimensional objects making them to “disappear”. The materials used for the invisibility devices
are artificially designed and research are being carried out since than the idea strike the mind of
thinkers of invisibility. The breakthrough could lead to systems for rendering anything from
people to large objects, such as tanks and ships, invisible to the eye – although this is still years
off. The day is not for away when we have invisibility cloak large enough to hide us.
Invisibility refers to the state of an object which cannot be seen. An object in this state is said to
be invisible (literally, "not visible"). The term is usually used as a fantasy/science fiction term,
where objects are literally made useable by magical or technological means. However, its effects
can also be seen in the real world, particularly in physics. Since objects can be seen by light in
the visible spectrum from a source reflecting off their surfaces and hitting the viewer's eye, the
most natural form of invisibility (whether real or fictional) is an object which does not reflect
light (that is - allows light to pass through it). In nature, this is known as transparency, and is
seen in many naturally-occurring materials (although no naturally-occurring material is 100%
transparent). The concept of a cloaking device, or the invisibility cloak, is to steer light waves
around an object to make it appear as if nothing were there. Visibility also depends on the eyes
of the observer and/or the instruments used. Thus an object can be classified as "invisible to" a
person, animal, instrument, etc.
The things which our eyes are not able to see are considered as “ Invisible”. Light is neither
absorbed nor reflected by the objects, passing like water flowing around a rock. As a result, only
the light from behind the objects can be seen. The devices which are going to make us hide are
invisibility devices. Researchers at the University of California at Berkeley, whose work is
funded by the American military, have engineered materials that can control light’s direction of
travel. The world’s two leading scientific journals, Science and Nature, are expected to report the
results in near future. . The concept of invisibility would involve surrounding the object by a
meta-material .

1
 Meta-material is a type of composite material that has unusual electromagnetic properties.
According to the researchers, light rays incident on the material would be bent around the object,
only to emerge on the other side in exactly the same direction as they began. Although the work
is only theoretical, the researchers reckon that materials invisible to radio waves could be
produced within five years.

1.2 Types of Cloaking

1.2.1Meta Material
1.2.2Optical Camouflage

Figure 1.1 Structure of a neural network

Neural networks are a set of algorithms, modeled loosely after the human brain, that are designed
to recognize patterns. They interpret sensory data through a kind of machine perception, labeling
or clustering raw input. The patterns they recognize are numerical, contained in vectors, into
which all real-world data, be it images, sound, text or time series, must be translated.
Input Layer – This refers to the input vector or the data that is being fed into the neural network
Hidden Layers – A hidden layer in an artificial neural network is a layer in between input layers
and output layers, where artificial neurons take in a set of weighted inputs and produce an output
through an activation function. It is a typical part of nearly any neural network in which
engineers simulate the types of activity that go on in the human brain.
Output Layer – This layer gives the predicted output of the network, we get our results from this
layer
Neural networks help us cluster and classify. We can think of them as a clustering and
classification layer on top of the data.They help to group unlabeled data according to similarities
among the example inputs, and they classify data when they have a labeled dataset to train on.
(Neural networks can also extract features that are fed to other algorithms for clustering and
classification; so you can think of deep neural networks as components of larger machine-

2
learning applications involving algorithms for reinforcement learning, classification
and regression.)
Some examples of the problems that we can solve using deep learning
1. Predicting the outcomes from the data : for example, spam or not_spam in an email
filter, good_guy or bad_guy in fraud detection, angry_customer or happy_customer in
customer relationship management.

2. Movie Recomendation systems on service like Amazon Prime Video ,Netflix ,etc. Music
recommendation systems on Spotify, computer vision etc.
1.2.1 Basics of Neural Networks

1) Neuron- Just like a neuron forms the basic element of our brain, a neuron forms the basic
structure of a neural network. Just think of what we do when we get new information.
When we get the information, we process it and then we generate an output. Similarly, in
case of a neural network, a neuron receives an input, processes it and generates an output
which is either sent to other neurons for further processing or it is the final output.

Figure 1.2 Struture of neuron

2) Weights – When input enters the neuron, it is multiplied by a weight. For example, if a
neuron has two inputs, then each input will have has an associated weight assigned to it.
We initialize the weights randomly and these weights are updated during the model
training process. The neural network after training assigns a higher weight to the input it
considers more important as compared to the ones which are considered less important. A
weight of zero denotes that the particular feature is insignificant.

Let’s assume the input to be a, and the weight associated to be W1. Then after passing
through the node the input becomes a*W1

3
 Figure1.3 Weights of neurons in a network

3) Bias – In addition to the weights, another linear component is applied to the input, called
as the bias. It is added to the result of weight multiplication to the input. The bias is
basically added to change the range of the weight multiplied input. After adding the bias,
the result would look like a*W1+bias. This is the final linear component of the input
transformation

4) Activation Function – Once the linear component is applied to the input, a non-linear
function is applied to it. This is done by applying the activation function to the linear
combination.The activation function translates the input signals to output signals. The
output after application of the activation function would look something like f(a*W1+b)
where f() is the activation function.

In the below diagram we have “n” inputs given as X1 to Xn and corresponding weights
Wk1 to Wkn. We have a bias given as bk. The weights are first multiplied to its
corresponding input and are then added together along with the bias. Let this be called as
u.

u=∑w*x+b -------- Equation (1)

The activation function is applied to u i.e. f(u) and we receive the final output from the
neuron as yk = f(u)

4
 Figure 1.4 Activation fucntion in a neural network

1.2.2 Applications of Neural Networks

1) Classification
All classification tasks depend upon labeled datasets; that is, humans must transfer their
knowledge to the dataset in order for a neural network to learn the correlation between labels and
data. This is known as supervised learning.
 Detect faces, identify people in images, recognize facial expressions (angry, joyful)
 Identify objects in images (stop signs, pedestrians, lane markers…)
 Recognize gestures in video
 Detect voices, identify speakers, transcribe speech to text, recognize sentiment in voices
 Classify text as spam (in emails), or fraudulent (in insurance claims); recognize sentiment
in text (customer feedback)

2) Clustering
Clustering or grouping is the detection of similarities. Deep learning does not require labels to
detect similarities. Learning without labels is called unsupervised learning. Unlabeled data is the
majority of data in the world. One law of machine learning is: the more data an algorithm can

5
train on, the more accurate it will be. Therefore, unsupervised learning has the potential to
produce highly accurate models.
 Search: Comparing documents, images or sounds to surface similar items.
 Anomaly detection: The flipside of detecting similarities is detecting anomalies, or
unusual behavior. In many cases, unusual behavior correlates highly with things you want
to detect and prevent, such as fraud.

3) Predictive Analytics: Regressions

Deep learning is able to establish correlations between, say, pixels in an image and the name of a
person. You might call this a static prediction. By the same token, exposed to enough of the right
data, deep learning is able to establish correlations between present events and future events. It
can run regression between the past and the future. The future event is like the label in a sense.
Deep learning doesn’t necessarily care about time, or the fact that something hasn’t happened
yet. Given a time series, deep learning may read a string of number and predict the number most
likely to occur next.
 Hardware breakdowns (data centers, manufacturing, transport)
 Health breakdowns (strokes, heart attacks based on vital stats and data from wearables)
 Customer churn (predicting the likelihood that a customer will leave, based on web
activity and metadata)
 Employee turnover (ditto, but for employees)

4) Text Classification and Categorization

Text classification is an essential part in many applications, such as web searching, information
filtering, language identification, readability assessment, and sentiment analysis. Neural
networks are actively used for these tasks.
 Used in Movie recommendation systems and other Customer systems
 Language Identification
 Statistical text categorization

5) Named Entity Recognition (NER)

The main task of named entity recognition (NER) is to classify named entities, such as Guido
van Rossum, Microsoft, London, etc., into predefined categories like persons, organizations,
locations, time, dates, and so on. Many NER systems were already created, and the best of them
use neural networks.
 Classifying content for News
 Efficient Search Algorithms
 Customer Reports

6
 6) Part-of-Speech Tagging
Part-of-speech (POS) tagging has many applications including parsing, text-to-speech
conversion, information extraction, and so on. In the work, Part-of-Speech Tagging with
Bidirectional Long Short-Term Memory Recurrent Neural Network a recurrent neural network
with word embedding for part-of-speech (POS) tagging task is presented.
 The model was tested on the Wall Street Journal data from Penn Treebank III data set
and achieved a performance of 97.40% tagging accuracy.
 Text to speech conversion

7) Semantic Parsing and Question Answering

Question Answering systems automatically answer different types of questions asked in natural
languages including definition questions, biographical questions, multilingual questions, and so
on. Neural networks usage makes it possible to develop high performing question answering
systems.
 The method uses the knowledge base at an early stage to prune the search space and thus
simplifies the semantic matching problem.
 It also applies an advanced entity linking system and a deep convolutional neural network
model that matches questions and predicate sequences. The model was tested on
WebQuestions data set, and it outperforms previous methods substantially.

8) Paraphrase Detection
Paraphrase detection determines whether two sentences have the same meaning. This task is
especially important for question answering systems since there are many ways to ask the same
question.
 It was shown that the proposed CNN model achieves high accuracy especially when the
words embedded are pre-trained on in-domain data. The model’s performance was
compared with Support Vector Machines and a duplicate detection approach. They
demonstrated that their CNN model outperforms the baselines by a large margin .

9) Language Generation and Multi-document Summarization

Natural language generation has many applications such as automated writing of reports,
generating texts based on analysis of retail sales data, summarizing electronic medical records,
producing textual weather forecasts from weather data, and even producing jokes.
 In a recent paper, Natural Language Generation, Paraphrasing and Summarization of
User Reviews with Recurrent Neural Networks, researchers describe a recurrent neural
network (RNN) model capable of generating novel sentences and document summaries.
 The paper described and evaluated a database of 820,000 consumer reviews in the
Russian language. The design of the network permits users control of the meaning of

7
 generated sentences. By choosing sentence-level features vector, it is possible to instruct
the network; for example, “Say something good about a screen and sound quality in
about ten words”.
 The ability of language generation allows production of abstractive summaries of
multiple user reviews that often have reasonable quality. Usually, the summary report
makes it possible for users to quickly obtain the information contained in a large cluster
of documents.

10) Machine Translation

Machine translation software is used around the world despite its limitations. In some domains,
the quality of translation is not good. To improve the results researchers try different techniques
and models, including the neural network approach.
 The purpose of Neural-based Machine Translation for Medical Text Domain study is to
inspect the effects of different training methods on a Polish-English machine translation
system used for medical data.
 To train neural and statistical network-based translation systems The European Medicines
Agency parallel text corpus was used.
 It was demonstrated that a neural network requires fewer resources for training and
maintenance. In addition, a neural network often substituted words with other words
occurring in a similar context .

Chapter - 2
Convolutional Neural Networks
2.1 Introduction
Similar to how we learn to recognise objects, we need to show an algorithm millions of pictures
before it is be able to generalize the input and make predictions for images it has never seen
before.
Computers ‘see’ in a different way than we do. Their world consists of only numbers. Every
image can be represented as 2-dimensional arrays of numbers, known as pixels.

8
 Figure 2.1 How computers see images

But the fact that they perceive images in a different way, doesn’t mean we can’t train them to
recognize patterns, like we do. We just have to think of what an image is in a different way.
To teach an algorithm how to recognise objects in images, we use a specific type of Artificial
Neural Network: a Convolutional Neural Network (CNN). Their name stems from one of the
most important operations in the network: convolution.

2.2 Architecture of Convolutional Neural Networks

A Convolutional Neural Network (CNN) is a Deep Learning algorithm which can take in an
input image, assign importance (learnable weights and biases) to various aspects/objects in the
image and be able to differentiate one from the other. The pre-processing required in a CNN is
much lower as compared to other classification algorithms. While in primitive methods filters are
hand-engineered, with enough training, CNN have the ability to learn these filters/characteristics.

9
 Figure 2.2 Structure of CNN

2.2.1 Input Image

In the figure, we have an RGB image which has been separated by its three color planes — Red,
Green, and Blue. There are a number of such color spaces in which images exist — Grayscale,
RGB, HSV, CMYK, etc.
Once the images reach dimensions, say 8K (7680×4320) it will get computationally
intensive.The role of the CNN is to reduce the images into a form which is easier to process,
without losing features which are critical for getting a good prediction.

Figure 2.3 How RGB imags are stored in computer

2.2.2 Convolution Layer

In the above demonstration, the green section resembles our 5x5x1 input image,as shown in fig
2.4.The element involved in carrying out the convolution operation in the first part of a
Convolutional Layer is called the Kernel/Filter, K, represented in the color yellow. We have
selected K as a 3x3x1 matrix.The similar process for RGB channel in shown in fig 2.5.

10
 Figure 2.4 Convolution operation for single channel image(ex-greyscale)

Figure 2.5 Convolution of RGB channel

2.2.3 Pooling Layer

Similar to the Convolutional Layer, the Pooling layer is responsible for reducing the spatial size
of the Convolved Feature. This is to decrease the computational power required to process the
data through dimensionality reduction. Furthermore, it is useful for extracting dominant features

11
which are rotational and positional invariant, thus maintaining the process of effectively training
of the model.
Pooling is broadly divided into 2 Classes: Max Pooling and Average Pooling.
 Max Pooling returns the maximum value from the portion of the image covered by the
Kernel.
 Average Pooling returns the average of all the values from the portion of the image
covered by the Kernel.
Max Pooling also performs as a Noise Suppressant. It discards the noisy activations altogether
and also performs de-noising along with dimensionality reduction. On the other hand, Average
Pooling simply performs dimensionality reduction as a noise suppressing mechanism. Hence, we
can say that Max Pooling performs a lot better than Average Pooling.

Figure 2.6 Types of Pooling

2.2.4 Activation Layer

12
It is used to determine the output of neural network like yes or no. It maps the resulting values in
between 0 to 1 or -1 to 1 etc. (depending upon the function).

The Activation Functions can be basically divided into 2 types-

1. Linear Activation Function

2. Non-linear Activation Functions

Activation layers work exactly as in other neural networks, a value is passed through a function
that squashes the value into a range. Some commonly used activation functions are shown in fig
2.7

Table 2.1 Commonly used Activation Functions

2.2.5 Classification — Fully Connected Layer (FC Layer)

13
The layer we call as FC layer, we flattened our matrix into vector (As shown in fig 2.9) and feed
it into a fully connected layer like neural network.
In the above diagram, feature map matrix will be converted as vector (x1, x2, x3, …). With the
fully connected layers, we combined these features together to create a model. Finally, we have an
activation function such as softmax or sigmoid to classify the outputs as cat, dog, car, truck etc.,

Figure 2.7 Fully connected layer

Figure 2.8 Flatening of an Imae matrix into a vector

2.3 Some CNN architectures

14
There are various architectures of CNNs available which have been key in building algorithms
which power and shall power AI as a whole in the foreseeable future. Some of them have been
listed below:
1. LeNet
2. AlexNet
3. VGGNet
4. GoogLeNet
5. ResNet
6. ZFNet

Chapter - 3
Conclusion and Future Scope

15
Convolutional Neural Networks are everywhere, they can find there application in virtually
every field. They can make the existing procedures and technologies more efficient and
powerful. They make theclassification and segmentation tasks easier as we dont have to
mannualy extract the features and train the model by ourselves.
The Neural networks are capable of finding the traits within in the data to find the most efficient
way to predict the outcomes. Convolutional neural network is a class of deep learning methods
which has become dominant in various computer vision tasks and is attracting interest across a
variety of domains, including radiology.Convolutional neural network is composed of multiple
building blocks, such as convolution layers, pooling layers, and fully connected layers, and is
designed to automatically and adaptively learn spatial hierarchies of features through a
backpropagation algorithm.

Future Scope
Integration of fuzzy logic into neural networks

 Fuzzy logic is a type of logic that recognizes more than simple true and false values,
hence better simulating the real world. For example, the statement today is sunny might
be 100% true if there are no clouds, 80% true if there are a few clouds, 50% true if it's
hazy, and 0% true if rains all day. Hence, it takes into account concepts like -usually,
somewhat, and sometimes.
 Fuzzy logic and neural networks have been integrated for uses as diverse as automotive
engineering, applicant screening for jobs, the control of a crane, and the monitoring of
glaucoma.

Pulsed neural networks

 Most practical applications of artificial neural networks are based on a computational

model involving the propagation of continuous variables from one processing unit to the
next. In recent years, data from neurobiological experiments have made it increasingly
clear that biological neural networks, which communicate through pulses, use the timing
of the pulses to transmit information and perform computation.

Hardware specialized for neural networks

 Some networks have been hardcoded into chips or analog devices , this technology will
become more useful as the networks we use become more complex.
 The primary benefit of directly encoding neural networks onto chips or specialized
analog devices is SPEED!
 NN hardware currently runs in a few niche areas, such as those areas where very high
performance is required in embedded applications of simple, hardwired networks (e.g.
voice recognition).
 Many NNs today use less than 100 neurons and only need occasional training. In these
situations, software simulation is usually found sufficient

16
  When NN algorithms develop to the point where useful things can be done with 1000's of
neurons and 10000's of synapses, high performance NN hardware will become essential
for practical operation.

Improvement of existing technologies

 All current NN technologies will most likely be vastly improved upon in the future.
Everything from handwriting and speech recognition to stock market prediction will
become more sophisticated as researchers develop better training methods and network
architectures.

NNs might, in the future, allow:

 Robots that can see, feel, and predict the world around them
 Improved stock prediction
 Common usage of self-driving cars
 Composition of music
 Handwritten documents to be automatically transformed into formatted word
processing documents
 Trends found in the human genome to aid in the understanding of the data
compiled by the Human Genome Project
 Self-diagnosis of medical problems using neural networks

17
 References
1. Tchoumatchenko,F. Vissotsky,J.-G. Ganascia “One approach to understand classification
by neural networks” , Proceedings of 1993 International Conference on Neural Networks
(IJCNN-93-Nagoya, Japan)
2. S.Albawi,T. A. Mohammed,S. Al-Zawi “Understanding of a convolutional neural
network” in Engineering and Technology (ICET),2017 International Conference on.
IEEE, 2017, pp. 1–6.

18