Chapter 9

Flashcards
ANNs are a type of machine learning model
that mimic the human brain to learn and make
decisions.

ANN are important for Big Data Handling,

Complex Pattern Recognition, and Flexibility in
various domains.

A type of machine learning model that mimic

the human brain to learn and make decisions.

A subset of machine learning that involves

using artificial neural networks (ANNs) with
multiple layers, also known as deep neural
networks.

Seeks to enable computers to interpret and

make decisions based on visual data, similarly
to how humans do.

Enables machines to 'see' and 'understand'

visual data, supports more informed and rapid
decision-making processes for individuals and
businesses.
 Manipulating images to achieve the desired
output, often improving the image's quality or
extracting certain features.

Goes beyond classification by identifying

objects in an image and providing their precise
location through bounding boxes.

Medical Diagnosis and Healthcare,

Environmental Monitoring, Agriculture,
Scientific Research, Robotics, and Augmented
Reality

Data Collection and Consent, Bias and

Discrimination, Deep fakes and Manipulation,
Personal Data Leaks

Technology that overlays digital information

into the real world, with computer vision
enabling devices to recognize and track
objects in the environment.

Allows machines to perceive and interact with

their environment in real-time, enabling them
to navigate complex environments and
manipulate objects with precision and
accuracy.

New potential use of computer vision might

Can you think of a new potential use of
become popular in the next decade, and some
computer vision that might become
benefits would be discussed such as its
popular in the next decade?
challenges and potential implications.
 Smart Notes
Introduction to Artificial Intelligence Lecture 9: Machine Learning Part(II): Artificial

Neural Networks, Deep Learning and Computer Vision

•Artificial Neural Networks (ANN)

•What are Artificial Neural Networks?

•Components of ANN

•Importance of Artificial Neural Networks

•Big Data Handling

•Complex Pattern Recognition

•Flexibility

•How do Artificial Neural Networks Work?

•Deep Learning: End-to-End Learning

•What is Deep Learning?

•Artificial Intelligence, Machine Learning & Deep Learning

•Machine learning Vs. Deep Learning

•Computer Vision (CV): Enabling a Machine to Interpret the Scene

•What is Computer Vision?

•Human Vision Vs Computer Vision

•Computer Vision Vs Image Processing

•Importance of Computer Vision in Today's World

•Computer Vision Primary Tasks

•Computer Vision Applications

•Ethics and Privacy in Computer Vision

Artificial Intelligence and Machine Learning

•Deep learning involves creating intelligent machines by learning patterns and

representations from data through a hierarchy of layers.

•Deep learning refers to neural networks with many hidden layers.

•AI is the broader concept of creating intelligent machines that mimic human behavior.

•Machine learning is a subset of AI, involving algorithms that learn from data to improve

performance on specific tasks.

•Deep learning uses deep neural networks to automatically learn complex patterns and

representations from data.

•Feature extraction is a significant step in traditional machine learning pipelines, involving

selecting or transforming important data elements to minimize dimensionality and highlight

task-relevant features.

•Deep learning automatically extracts features from raw input, reducing the need for

extensive manual feature engineering in many cases.

•Computer vision is a branch of AI enabling machines to interpret and understand visual

information from the world around us.

•The scene refers to the environment or object being observed, captured by the eyes and
processed by the brain in human vision, and by a sensing device such as a camera or

scanner in computer vision.

•Images and videos are used in computer vision to enable machines to accurately identify

and classify objects.

Navigating the Breadth of Visual Data Types

•Thermal Images: Captured using infrared cameras, these images represent temperature

differences in a scene.

•Satellite Image: High-resolution images captured by satellites, often used for mapping,

agriculture, and environmental monitoring.

•Medical Image: Specialized images from devices like X-rays, MRIs, CT scans, and

ultrasounds.

•Grayscale Images: Black and white or monochromatic images.

•Recorded Videos: Stored video files from various sources.

•Real-time Streams: Such as feeds from CCTV or webcams.

•Sensing Device: In computer vision, the sensing device is typically a camera or a sensor

array that captures visual information from the environment.

•Computer Vision: Once the image/ video is captured, it undergoes various processing

stages, depending on the task.

•Image Processing: It deals with manipulating images to achieve the desired output, often

improving the image's quality or extracting certain features.

•The Surge of Visual Content: Given the vast amounts of visual content, there's an urgent
need for AI algorithms to automatically analyze and interpret these videos, a task beyond

human capacity.

•Facilitating Real-time Processing: Computer vision allows for instantaneous analysis,

ensuring timely reactions to dynamic situations.

•Automation and Efficiency: By enabling machines to 'see' and 'understand' visual data,
computer vision supports more informed and rapid decision-making processes for

individuals and businesses.

•The Importance of Computer Vision in Today's World: Detection, Recognition, Pose
Estimation, Object detection, Tracking, and Segmentation are primary tasks of computer

vision. This involves training a model to categorize images into predefined classes or
categories, such as classifying whether an image contains a cat or a dog. Object detection

goes beyond classification by identifying objects in an image and providing their precise

location.

Computer Vision Primary Tasks

•Image segmentation involves dividing an image into segments or regions that correspond

to meaningful parts of the image, aiming to partition it into regions with similar properties.

•Applications of segmentation include medical imaging to identify and isolate specific

structures or abnormalities, satellite and remote sensing to analyze and classify land cover,

and image editing for background removal and object extraction.

•Object tracking involves following the movement of an object across a sequence of images

or frames in a video.

•Uses of object tracking include road condition monitoring, autonomous vehicles for

collision avoidance, sports analysis, augmented reality, and robotics.

•Pose estimation aims to determine the position and orientation of an object or person in

an image or video frame, involving identifying key points and estimating their positions.

•Applications of pose estimation include virtual reality, gesture recognition, and healthcare

for monitoring body movements and posture.

•Computer vision tasks with potential societal impact in the next decade include security
applications such as facial recognition, object detection, and anomaly detection for

surveillance systems, access control, and fraud detection in the automotive industry, retail
and e-commerce for inventory management, customer analytics, personalized shopping

recommendations, and visual search engines, and accessibility.

Applications and Implications of Computer Vision

•It can improve the accessibility of digital content for individuals with disabilities.

•It aids in medical image analysis, helping doctors detect diseases, tumors, and anomalies in
X-rays, MRIs, and CT scans. It can also be used in monitoring patient movement and vital

signs.

•It can be employed in monitoring and analyzing satellite and drone imagery for
environmental purposes, such as tracking deforestation, studying climate change, and

assessing wildlife populations.

•Precision agriculture uses computer vision to monitor crop health, detect diseases, and

optimize resource allocation, leading to higher crop yields and more sustainable farming

practices.

•Computer vision is used in various scientific disciplines to analyze and interpret visual data,

aiding researchers in their studies.

•In robotics, computer vision allows machines to perceive and interact with their
environment in real-time, enabling them to navigate complex environments, manipulate

objects with precision, and assist in search and rescue missions.

•Augmented reality (AR) utilizes computer vision by enabling devices to recognize and track
objects in the environment, allowing AR applications to integrate virtual content with the

physical world seamlessly.

•With advancing technology, a potential new use of computer vision might be in personal
healthcare monitoring. Leveraging computer vision to track and analyze individual health
and movement patterns could provide valuable insights for preventive medicine and early
detection of health issues. The benefits could include personalized and proactive healthcare,
enhanced patient engagement, and improved health outcomes. However, challenges may

include privacy concerns, data security, and the ethical use of personal health information.
•As computer vision technology becomes more advanced and widespread, it is important to
consider the ethical and privacy concerns that come with it, such as data collection and

consent, bias and discrimination, deep fakes and manipulation, and personal data leaks.

Ethical and Privacy Dilemmas of Computer Vision

•Computer vision can pose ethical dilemmas in scenarios such as using facial recognition

technology in public spaces, which can infringe on individual privacy rights.

•The potential benefits of computer vision technology include advancements in security and

law enforcement, as well as improved efficiency in industries like retail and healthcare.

•However, the risks involve invasion of privacy, potential misuse of data, and the loss of

anonymity in public spaces, leading to increased surveillance and tracking of individuals.

•Balancing technological advancements with preserving individual rights and privacy

requires implementing strict regulations and ethical guidelines for the use of computer
vision technology, as well as ensuring transparency and accountability in decision-making

processes.
 Summary
"The Intersection of ANNs, Deep Learning, and Computer Vision"

This lecture covers Artificial Neural Networks (ANNs), Deep Learning, and Computer Vision.

ANNs are machine learning models that mimic the human brain to learn and make
decisions, handling big data, recognizing complex patterns, and being flexible in various
applications. Deep Learning is a subset of machine learning that uses ANNs with multiple
layers. Computer Vision enables machines to interpret scenes, with applications in image

and video processing, and is essential in today's world.

"Unlocking Insights: The Intersection of Deep Learning and Computer Vision"

Deep learning is a subset of machine learning using deep neural networks to automatically
learn complex patterns and representations from data. It reduces the need for manual
feature engineering. Computer vision is a branch of AI that enables machines to interpret
and understand visual information from the world around us, using algorithms and

techniques to analyze, process, and make sense of images and videos.

"Empowering Automation with Computer Vision: The Role of AI in Analyzing Visual

Data"

Computer Vision involves processing visual data from various sources, such as images,
videos, and medical scans. It aims to enable machines to interpret and make decisions

based on visual information, similar to human perception. Primary tasks include image
classification, object detection, face recognition, optical character recognition, and

segmentation. The surge of visual content highlights the need for AI algorithms to
automatically analyze and interpret this data, enabling automation, efficiency, and real-time

processing for informed decision-making.

The Future of Computer Vision: Enhancing Societal Impact through Deep Learning

Computer vision has various tasks like image classification, object detection, pose
estimation, and image segmentation. It holds potential for societal impact in areas such as

security, automotive industry, retail, and accessibility. In the next decade, deep learning
advancements may further enhance these tasks, leading to more efficient autonomous

driving, better retail experiences, improved accessibility for visually impaired individuals, and

advanced security systems.

The Diverse Applications and Ethical Concerns of Computer Vision Technology

Computer vision technology has numerous applications, including accessibility for

individuals with disabilities, medical diagnosis, environmental monitoring, precision
agriculture, scientific research, robotics, and augmented reality. Its potential future use could
be in personalized advertising, where computer vision analyzes a user's environment and
preferences to display relevant ads. This could improve user experience but may raise

concerns about privacy and data misuse.

"Ethical and Privacy Considerations in Computer Vision Technology"

Computer vision may pose ethical and privacy dilemmas, such as loss of anonymity and
over-reliance, in scenarios like automated decision-making. While the technology offers

benefits like improved efficiency, the risks include a lack of accountability and potential
privacy breaches. To balance advancements with individual rights, it's crucial to implement

measures that protect users' data and ensure transparency in AI decision-making processes.
Quiz
‫‪Good luck on your finals J‬‬

‫‪Made by: Rayan MohammadNaeem Tashkandy‬‬

‫ﻻ ﺗ ﻧ ﺳ و ﻧ ﺎ ﻣ ن ﺻ ﺎ ﻟﺢ د ﻋ ﺎ ﺋ ﻛ م ﷲ ﯾ و ﻓ ﻘ ﻛ م ﯾ ﺎ ر ب ا ﻣ ﯾ ن‬
‫ر اﺟ ﻌ و ا اﻟﺳ ﻼ ﯾد ا ت ﻛ ﻣ ﺎ ن ا ﺣ ﺗ ﯾ ﺎ ط‬