Google AI-ML Virtual Internship

Internship Report submitted in partial fulfillment of the

requirements for the award of the degree of

BACHELOR OF TECHNOLOGY

IN

COMPUTER SCIENCE AND ENGINEERING

By

Y. Vijaya Chandra
22R11A05J4

Department of Computer Science and Engineering

Accredited by NBA

Geethanjali College of Engineering and Technology

(UGC Autonomous)
(Affiliated to J.N.T.U.H, Approved by AICTE,
New Delhi) Cheeryal (V), Keesara (M),
Medchal.Dist.-501 301.

August -2024
 Geethanjali College of Engineering and
Technology
(UGC Autonomous)
(Affiliated to JNTUH, Approved by AICTE, New Delhi)
Cheeryal (V), Keesara(M), Medchal Dist.-501 301.

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

Accredited by NBA

CERTIFICATE

This is to certify that the Internship Report entitled “Google AI-ML Virtual
Internship” is a bonafide work done by Y. Vijaya Chandra ,22R11A05J4 in partial
fulfillment of the requirements for the award of the degree of Bachelor of Technology
in “Computer Science and Engineering” from Jawaharlal Nehru Technological
University, Hyderabad during the year 2024-2025.

HOD - CSE
Dr. A Sree Lakshmi,
Professor

Examiner
Signature:
Name:
Designation:
 Geethanjali College of Engineering and
Technology
(UGC Autonomous)
(Affiliated to JNTUH Approved by AICTE, New Delhi)
Cheeryal (V), Keesara(M), Medchal Dist.-501 301.

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

Accredited by NBA

DECLARATION BY THE CANDIDATE

I, Y. Vijaya Chandra, bearing Roll No .22R11A05J4 hereby declare that the Internship

Report entitled “Google AI-ML Virtual Internship” is submitted in partial fulfillment

of the requirements for the award of the degree of Bachelor of Technology in Computer

Science and Engineering.

This is a record of bonafide work carried out by me in E D U S K I L L S and the

results embodied in this internship report have not been reproduced or copied from
any source. The results embodied in this Internship Report have not been submitted
to any other University or Institute for the award of any other degree or diploma.

Y. Vijaya Chandra,

22R11A05J4,

CSE,
GCET
 ACKNOWLEDGEMENT

First and foremost, I would like to extend my heartfelt gratitude to Edu Skills, with Google
serving as the invaluable bridge that facilitated this enriching internship within the organization.
I am profoundly indebted to Director Mr. G.R. Ravinder Reddy and our esteemed
Principal Dr. S. Udaya Kumar Susarla for their unwavering support and for providing the
necessary facilities that have enabled me to make the most of this internship.
I extend my deepest thanks to my Head of the Department, Dr. A Sree Lakshmi, for
her invaluable insights and constructive feedback that have been instrumental in shaping my
internship experience.
I am immensely appreciative of the dedicated faculty members and staff within the
department, whose consistent guidance and support have been pivotal in my growth during this
internship.
Furthermore, I would like to express my sincere gratitude for the exceptional skill
development opportunities that Edu Skills, facilitated through Google, has provided me with.
The hands-on experience and practical knowledge gained have been instrumental in augmenting
my professional skill set, preparing me comprehensively for my future pursuits.

Y. Vijaya Chandra,

22R11A05J4,
 INTRODUCTION ABOUT INTERNSHIP ORGANIZATION

Edu Skills is a non-profit organization committed to developing an Industry 4.0 -ready digital
workforce in India. With a vision to bridge the gap between academia and industry, Edu Skills
ensures that faculties and students have access to world-class curricula. We aim to transform the
teaching methodologies and ICT-based education systems in India by working closely with key
stakeholders, including students, faculties, educational institutions, and government bodies. Our
mission is to positively impact 1 million beneficiaries by 2024 through targeted skill interventions.
By identifying skill gaps in students and aligning them with the latest global technical skills, we
provide innovative learning aids and services that benefit the entire education ecosystem. Our
values focus on fostering connections between industry-ready professionals, researchers,
educators, and entrepreneurs, and extending cutting-edge research and consultancy to advance
global knowledge and skills for the next generation. Edu Skills' holistic approach addresses
education, employment, and entrepreneurship, driving social and business impact across India.
 TRAINING SCHEDULE

WEEK DATE WORKDONE

NO.
Week (1-2) 1 April – 14 April Briefing about Artificial Intelligence
and Machine Learning
Week (3-4) 15 April – 28 April Introduction of Python Language
Week 5 29 April – 5 May Program neural networks with
TensorFlow
Week 6 6 May – 12 May Get started with object detection
Week 7 13 May – 19 May Go further with object detection
Week 8 20 May – 26 May Get started with product image search
Week 9 27 May – 2 June Go further with product image search
Week 10 3 June – 9 June Go further with image classification
 TABLE OF CONTENTS

S.NO. CONTENT PAGE NO.

a Abstract i

b List of Figures ii

c List of Symbols & Abbreviations iii

1 Introduction 1

2 Technologies used 4

3 Work done 6

4 Learning after Internship 11

5 Conclusion 13

6 Bibliography 14
 ABSTRACT

AI-ML Virtual Internship

April - June | Hosted by Edu skills

This 10-week AI-ML Virtual Internship offers a thorough introduction to the fundamentals of
Artificial Intelligence and Machine Learning. The program starts with a deep dive into core AI
and ML concepts, followed by Python programming essentials tailored specifically for these
fields. As participants progress through the internship, they engage in data preprocessing,
explore supervised and unsupervised learning models, and complete hands-on projects to
solidify their understanding.

Advanced topics such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks
(RNNs) are covered in depth, with a focus on their practical applications in real-world
scenarios. The program also addresses the ethical dimensions of AI, exploring issues like bias
and the broader societal impacts of AI technologies.

In the final weeks, participants undertake a capstone project, applying the skills and knowledge
they've acquired to solve a real-world problem. This project serves as the internship's culmination,
showcasing each participant's ability to develop and present a comprehensive AI/ML solution.
Upon completion, participants will have gained the essential skills and experience needed to
pursue a career or further studies in AI and ML.

i
 LIST OF FIGURES

S.NO. FIG.NO. TITLE OF FIGURE PAGE.NO.

1. 1.2 Importance of AI 2
2. 1.3 Applications of AI 3
 LIST OF ABBREVIATIONS

ABBREVIATIONS EXPANSION

AI Artificial Intelligence
ML Machine Learning
IDE Integrated Development Environment
PY Python
RAM Random Access Memory
CNN Convolutional Neural Networks
CPU Computer Processing Unit
 1. INTRODUCTION

1.1 DEFINITION OF ARTIFICIAL INTELLIGENCE

Artificial Intelligence (AI) stands at the forefront of computer science, with a primary focus on
empowering machines to execute tasks traditionally reserved for human intelligence. These
tasks encompass learning, reasoning, problem-solving, perception, and comprehending natural
language. AI endeavors to emulate human-like thinking, learning, and decision-making by
employing advanced algorithms, data-driven methods, and intricate computational models.
Noteworthy subfields within AI encompass machine learning, natural language processing,
computer vision, robotics, among others.

1.2 SIGNIFICANCE OF AI
In recent years, AI has emerged as a transformative force, reshaping approaches to
complex problems and unlocking unprecedented possibilities across various industries. Its
applications traverse diverse domains, including healthcare, finance, manufacturing,
transportation, and entertainment. The paramount significance of AI lies in its ability to process
extensive volumes of data and extract meaningful insights, empowering businesses to make
well- informed decisions with unparalleled accuracy. Additionally, AI-powered systems have
showcased remarkable proficiency in tasks such as image and speech recognition, language
translation, autonomous navigation, and personalized recommendation systems.
 Fig.No.1.2 Importance of AI

1.3 APPLICATIONS OF AI
In recent years, Artificial Intelligence (AI) has emerged as a transformative force across
various industries, fundamentally reshaping how we approach complex challenges and
unlocking unprecedented potentials. This cutting-edge technology empowers machines to
replicate human- like cognitive functions, including learning, reasoning, problem-solving,
perception, and understanding natural language. Through the utilization of advanced algorithms,
data-driven techniques, and sophisticated computational models, AI is positioned to
revolutionize our interaction with technology and enhance human capabilities.
Healthcare: AI enhances medical diagnostics through precise image analysis and expedites drug
discovery by predicting potential candidates based on biological data.
Finance: AI optimizes investment strategies through algorithmic trading and strengthens fraud
prevention by identifying irregular transaction patterns.
Manufacturing: AI predicts machinery failures for proactive maintenance and ensures higher
quality through product inspections in manufacturing.
Autonomous Vehicles: AI enables self-driving cars for improved road safety and traffic
management through real-time data analysis.
Education: AI tailors educational content based on individual learning patterns and provides
personalized tutoring for understanding complex concepts.
Cybersecurity: AI identifies and mitigates cybersecurity threats by analyzing patterns and
anomalies in network traffic and user behavior.

Fig.No.1.3 Applications of AI
 2. TECHNOLOGIES USED

Convolutional Neural Networks (CNNs):

CNNs are a type of deep learning architecture particularly well-suited for image
recognition tasks. They have been successfully applied to analyze medical images, including chest
X-rays, for pneumonia detection.

Python:
Python is a versatile programming language with a broad range of applications, and
various technologies are used in conjunction with Python to enhance its capabilities. Python has
a vast standard library and a large community, making it widely used for web development, data
science, artificial intelligence, and automation. Its syntax emphasizes code readability and ease of
use, making it an excellent choice for both beginners and experienced developers.

2.1 Hardware Requirements

 CPU: A modern multi-core CPU is essential. While a powerful CPU will help speed up
training, even mid-range CPUs with multiple cores can handle the task. Minimum
(intel i5 or AMD Ryzen 5) of latest generations is recommended.
 RAM: A minimum of 8GB of RAM is recommended. More RAM is beneficial, especially
when working with larger datasets. If your dataset is small to moderate in size, 8GB should
be sufficient.
 GPU (Optional): While not explicitly mentioned in your script, having a compatible
GPU (NVIDIA GPU, for example) can significantly accelerate deep learning tasks.
TensorFlow can be configured to use GPU acceleration, which speeds up the training
process. However, the integrated GPU provided i.e. (iris xe graphics of intel, Radeon
graphics of AMD) will be sufficient
 Storage: Ensure that you have enough storage space for the dataset and the saved model.
Deep learning models and datasets can be large, so having ample storage is crucial. SSD
type of storage is recommended.
2.2 Software Requirements

 Operating System: The program is designed to run on operating systems that

support Python and TensorFlow. Commonly used operating systems include:
Windows 10, Linux (Ubuntu, CentOS, etc.) macOS.
 Python: Python is the primary programming language used in the script. Ensure
that Python is installed on the system. The code was written using Python 3.10.
You can download Python from the official Python website.
 Python Libraries: The program utilizes several Python libraries that need to be
installed. Libraries like tensors, matplotlib, keras etc. are required.
 Development Environments: If you are planning to modify or extend the code,
an integrated development environment (IDE) is recommended. Popular Python
IDEs include: PyCharm, Visual Studio Code, Jupyter Notebooks (for interactive
development)
 3. WORK DONE

3.1 Prerequisite
As a prerequisite for this internship, we engaged in an intensive Python training program
conducted from April 1 to April 15, 2024. This e-learning phase imparted fundamental
programming skills, encompassing vital concepts such as version control, functions, and data
manipulation. This training laid a sturdy foundation for advanced modules and is key to our
readiness for the challenges that lie ahead in this internship.

3.2 AI-ML course units covered

UNIT-1 Program neural networks with
TensorFlow UNIT-2 Get started with object
detection
UNIT-3 Go further with object detection
UNIT-4 Get started with product image
search UNIT-5 Go further with product image
search UNIT-6 Go further with image
classification

UNIT-1 Program neural networks with TensorFlow

The "Program Neural Networks with TensorFlow" unit in the Google AI-ML course introduces
the fundamentals of building and training neural networks using TensorFlow, a powerful open-
source machine learning framework. The unit begins with an overview of neural networks,
explaining key concepts such as layers, activation functions, and the flow of data through a
network. Participants learn how to define and structure neural networks in TensorFlow, using its
intuitive APIs like Keras to create models that can handle various tasks, from basic image
classification to more complex functions. The unit also covers the process of compiling and
training the model, introducing concepts like loss functions, optimizers, and metrics that are
crucial for evaluating the performance of a neural network.
In addition to theoretical knowledge, this unit emphasizes hands-on practice by guiding learners
through building and training their own neural networks. By experimenting with different
architectures and hyperparameters, participants gain a deeper understanding of how to optimize
and improve model performance. The unit concludes with techniques for validating and testing
models to ensure they generalize well to new data, as well as an introduction to saving and loading
trained models for future use. This foundational unit equips learners with the essential skills
needed to leverage TensorFlow for a wide range of AI and machine learning applications.

UNIT-2 Get started with object detection

"Get Started with Object Detection" focuses on the techniques and tools required to identify and
locate objects within images or video frames. This unit introduces fundamental object detection
concepts, including bounding boxes, class labels, and the use of convolutional neural networks
(CNNs) for feature extraction. Learners become familiar with popular object detection models,
such as YOLO (You Only Look Once) and SSD (Single Shot MultiBox Detector), and explore
how these models can be applied to detect multiple objects in an image with varying sizes and
orientations. The unit emphasizes the importance of dataset preparation and annotations,
guiding participants through the process of preparing training data and evaluating model
performance.

Hands-on exercises are a key component, allowing learners to implement and fine-tune object
detection models using frameworks like TensorFlow or PyTorch. By working with real-world
datasets and experimenting with different hyperparameters and model architectures, participants
gain practical experience in improving detection accuracy and handling challenges such as
occlusions and varying object scales. The unit equips learners with the skills needed to build
robust object detection systems that can be applied to various applications, from security
surveillance to autonomous vehicles.
UNIT-3 Go further with object detection

"Go Further with Object Detection" delves into advanced techniques and improvements for
enhancing the performance of object detection systems. Building on foundational concepts, this
unit explores sophisticated models and algorithms that offer more accurate and efficient detection.
Topics may include region-based approaches like Faster R-CNN, which refines object detection
by proposing regions of interest before classification, and multi-scale detection methods that
improve performance across different object sizes. The unit also covers techniques for handling
challenging scenarios such as overlapping objects, varying lighting conditions, and complex
backgrounds.

In addition to model enhancements, learners gain insight into optimizing detection systems for
real-world applications. This includes strategies for deploying models in production
environments, improving inference speed, and integrating object detection with other computer
vision tasks. Practical exercises involve experimenting with advanced architectures, fine-tuning
pre-trained models, and employing data augmentation techniques to further boost accuracy. The
unit aims to equip learners with a deeper understanding of object detection and the skills needed
to tackle more complex challenges and deploy robust solutions in diverse scenarios.

UNIT-4 Get started with product image search

"Get Started with Product Image Search" introduces the techniques and tools used to build systems
that allow users to search for products based on images rather than text. The unit begins by
explaining the fundamentals of image retrieval and similarity search, emphasizing the importance
of feature extraction and representation. Learners are introduced to methods such as extracting
visual features from images using convolutional neural networks (CNNs) and leveraging these
features to compare and rank images in a database. This enables users to input an image and
retrieve visually similar products, facilitating a more intuitive and efficient search experience.
The unit also covers practical aspects of implementing a product image search system, including
dataset preparation, image indexing, and query handling. Participants gain hands-on experience
by building and evaluating their own image search systems using frameworks and libraries that
support feature extraction and similarity computation. By experimenting with different techniques
and optimizations, learners develop the skills needed to create a functional and accurate image
search application, addressing challenges such as scaling, relevance ranking, and user experience.

UNIT-5 Go further with product image search

"Go Further with Product Image Search" builds on the foundational concepts introduced in the
previous unit by exploring advanced techniques and optimizations for enhancing product image
search systems. This unit delves into sophisticated methods for improving search accuracy and
efficiency, such as leveraging deep learning-based feature embeddings and advanced similarity
metrics. Learners explore techniques for refining image representations, such as using transfer
learning with pre-trained models or implementing custom architectures tailored to specific
product categories. Additionally, the unit covers approaches for handling large-scale image
databases, including distributed indexing and retrieval systems to ensure fast and scalable
search performance.

The unit also addresses real-world challenges and practical considerations in deploying image
search systems. This includes techniques for handling diverse and high-dimensional data,
optimizing search algorithms for both relevance and speed, and integrating feedback
mechanisms to improve search results over time. Learners gain hands-on experience with
advanced tools and frameworks, applying these techniques to build and fine-tune a robust
product image search system. By focusing on these advanced aspects, the unit equips learners
with the knowledge and skills needed to implement sophisticated image search solutions in
production environments.
UNIT-6 Go further with image classification

"Go Further with Image Classification" extends the foundational knowledge of image
classification by exploring more advanced techniques and methodologies to enhance model
performance and adapt to complex scenarios. This unit delves into deep learning architectures
such as convolutional neural networks (CNNs), including advanced variants like ResNet,
Inception, and DenseNet, which address challenges like vanishing gradients and improved feature
extraction. Learners gain insight into techniques for fine-tuning these models on specialized
datasets, leveraging transfer learning to adapt pre-trained models for new classification tasks,
and implementing sophisticated regularization methods to prevent overfitting and improve
generalization.

Additionally, the unit covers strategies for addressing real-world challenges in image
classification, such as handling class imbalance, improving model robustness against
adversarial attacks, and deploying models in resource-constrained environments. Practical
exercises involve experimenting with data augmentation, hyperparameter tuning, and evaluating
model performance using advanced metrics and validation techniques. By integrating these
advanced concepts and tools, the unit aims to equip learners with the expertise needed to tackle
complex image classification problems and develop high-performing models for diverse
applications.
 4. LEARNING AFTER INTERNSHIP

1. Enhanced Understanding of AI and ML

o Deepened my knowledge of various AI and ML techniques, including supervised
and unsupervised learning, neural networks, and deep learning models.
o Gained a more refined understanding of how different algorithms and models are
applied to address real-world challenges.

2. Practical Application of Theoretical Concepts

o Bridged the gap between theory and practice by learning to implement and fine-tune
models using real-world datasets.
o Acquired hands-on experience with tools like Python, TensorFlow, Keras, and
Scikit- Learn, applying them effectively in various projects.

3. Data Handling and Preprocessing Skills

o Enhanced my abilities in data cleaning, preprocessing, and visualization, essential
for preparing data for analysis and model training.
o Gained experience in tackling data challenges, such as handling missing values,
normalization, and feature extraction.

4. Model Development and Evaluation

o Developed and evaluated diverse machine learning models, learning the significance
of performance metrics and optimizing hyperparameters.
o Utilized tools for model evaluation, including cross-validation and performance
metrics, to ensure the accuracy and robustness of models.
5. Collaboration and Communication
o Improved my ability to work collaboratively with peers and mentors, sharing
insights, reviewing code, and actively contributing to group discussions.
o Enhanced my skills in documenting work, preparing reports, and presenting findings
clearly and effectively.

6. Ethical Considerations in AI
o Cultivated an awareness of the ethical implications of AI technologies, particularly
concerning bias, fairness, and the societal impact of AI solutions.

7. Project Management
o Gained experience in managing end-to-end projects, from initial data collection and
model development to final deployment and presentation of results.
o Learned to manage project timelines, deliverables, and integrate various components
to achieve project goals.
Overall, this internship provided invaluable practical experiences, deepening my understanding
of AI and ML concepts, while also enhancing the technical and soft skills essential for a career
in the field.
 5. CONCLUSION

The 10-week AI-ML Virtual Internship, held from April 1st to June 9th, provided a
comprehensive exploration of the field of Artificial Intelligence and Machine Learning. This
program seamlessly integrated theoretical knowledge with practical application, equipping me
with essential skills in data preprocessing, model development, and advanced deep learning
techniques.

Key achievements included mastering critical technologies, successfully executing hands-on

projects, and enhancing both technical and soft skills. The capstone project, in particular,
highlighted my ability to develop and deploy effective AI solutions, demonstrating real-world
applicability. Additionally, the internship deepened my understanding of the ethical
considerations surrounding AI, preparing me for responsible and impactful work in the field.

Overall, this internship significantly contributed to my professional growth and laid a strong
foundation for pursuing further studies or career opportunities in AI and ML.
 6. BIBLIOGRAPHY

1. https://www.ibm.com/topics/artificial-intelligence
2. https://www.w3schools.com/python/
3. https://www.tutorialspoint.com/python/index.htm
4. https://www.geeksforgeeks.org/artificial-intelligence-an-introduction/
5. https://www.javatpoint.com/artificial-intelligence-ai
6. https://www.geeksforgeeks.org/artificial-neural-networks-and-its-applications/