Future Trends in Artificial Intelligence
Dr. U. Srinivasulu Reddy
Associate Professor
Associate Dean (Students’ Welfare)
Department of Computer Applications
National Institute of Technology
Tiruchirappalli
29-08-2024 [email protected] 1
Mimic of Five Senses
29-08-2024 [email protected]
2
Evolution of Artificial Intelligence
29-08-2024 [email protected] 3
What is Artificial Intelligence (AI)
❖ What is Artificial Intelligence?
• It is the science and engineering of making intelligent machines,
especially intelligent computer programs
• It is related to the similar task of using computers to understand human
intelligence
❖ What is intelligence?
• Intelligence is the computational part of the ability to achieve goals in the
world
• “the capacity to learn and solve problems” (Webster's
dictionary)
• in particular
• the ability to solve novel problems
• the ability to act rationally
• the ability to act like humans
Artificial Intelligence build and understand intelligent entities or agents
2 main approaches: “engineering” versus “cognitive modelling”
29-08-2024 [email protected] 4
Artificial Intelligence Techniques
Brain
Artificial Neural Network(ANN)
Evolutionary Theory
Genetic Algorithms (GA)
29-08-2024
[email protected] 5
Artificial Intelligence Techniques
Birds
Fishes
Particle Swarm Optimization (PSO)
Ants
Ant Colony Optimization (ACO)
[email protected] 29-08-2024 6
Artificial Intelligence Techniques
Human
Fuzzy Systems & Fuzzy Logic (FS & FL)
29-08-2024 [email protected] 7
Evolution of AI — Source: https://www.embedded-vision.com/
29-08-2024 [email protected] 8
Machine Learning
9
[email protected]
What is Learning
• “Learning denotes changes in a system that ... enable a system to do
the same task … more efficiently the next time.” - Herbert Simon
• “Learning is constructing or modifying representations of what is
being experienced.” - Ryszard Michalski
• “Learning is making useful changes in our minds.” - Marvin Minsky
“Machine learning refers to a system capable of the autonomous
acquisition and integration of knowledge.”
29-08-2024 [email protected] 10
What is Machine Learning?
In 1959, Arthur Samuel defined machine learning as a "Field of study that
gives computers the ability to learn without being explicitly programmed".
Tom M. Mitchell provided a more formal definition
“A computer program is said to learn from experience E with respect to some
class of tasks T and performance measure P, if its performance at tasks in T, as
measured by P, improves with experience E”
Assume we want to classify an incoming mail as a Spam or Not. This problem
can be described in terms of three elements as below :
Spam Mail detection learning problem
Task T: To recognize and classify emails into ‘spam’ or ‘not spam’.
Performance measure P: Total percent of mails being correctly classified as
‘spam’ (or ‘not spam’ ) by the program.
Training experience E: A set of mails with given labels (‘spam’ / ‘not spam’).
Machine Learning
11
[email protected] What is Machine Learning?
• “Machine Learning at its most basic is the practice of using algorithms to parse
data, learn from it, and then make a determination or prediction about
something in the world.” – Nvidia
• “Machine learning is the science of getting computers to act without being
explicitly programmed.” – Stanford
• “Machine learning is based on algorithms that can learn from data without
relying on rules-based programming.”- McKinsey & Co.
• “Machine learning algorithms can figure out how to perform important tasks by
generalizing from examples.” – University of Washington
• “The field of Machine Learning seeks to answer the question “How can we
build computer systems that automatically improve with experience, and what
are the fundamental laws that govern all learning processes?” – Carnegie
Mellon University
Machine Learning
12
[email protected]
Traditional Programming
Data
Computer Output
Program
Machine Learning
Data
Computer Program
Output
29-08-2024 [email protected] 13
Related Fields
data
mining control theory
statistics
decision theory
information theory Machine
Learning
cognitive science
databases
psychological models
evolutionary neuroscience
models
Machine learning is primarily concerned with the accuracy and
effectiveness of the computer system.
Machine Learning
14 [email protected]
Machine
Learning
Data Science
£££
Software
Statistician
Engineer
Specialist
Domain
Knowledge
29-08-2024 [email protected] 15
When Do We Use Machine Learning?
• Human expertise does not exist (navigating on Mars)
• Humans can’t explain their expertise (speech recognition)
• Models must be customized (personalized medicine)
• Models are based on huge amounts of data (genomics)
29-08-2024 [email protected] 16
29-08-2024 [email protected] 17
Machine Learning Algorithms
Supervised learning Unsupervised learning
Semi-supervised learning
29-08-2024 [email protected] 18
Unsupervised Learning
29-08-2024 [email protected] 19
15
Reinforcement Learning
Supervised
Learning
29-08-2024 [email protected] 20
Smart Home
29-08-2024 [email protected] 21
29-08-2024 [email protected] 22
29-08-2024 [email protected] 23
29-08-2024 [email protected] 24
29-08-2024 [email protected] 25
29-08-2024 [email protected] 26
29-08-2024 [email protected] 27
29-08-2024 [email protected] 28
29-08-2024 [email protected] 29
Using machine learning to recommend books
ALGORITHMS
Collaborative Filtering
Nearest Neighbour
29-08-2024 ML to DL [email protected] Clustering 30
Using machine learning to identify faces and expressions
ALGORITHMS
Decision Trees
Adaboost
29-08-2024 [email protected] 31
Using machine learning to identify vocal patterns
ALGORITHMS
Feature Extraction
Probabilistic Classifiers
Support Vector Machines
+ many more….
29-08-2024 [email protected] 32
Machine Learning for working with social network data:
detecting fraud, predicting click-thru patterns, targeted advertising, etc.,
ALGORITHMS
Support Vector Machines
Collaborative filtering
Rule mining algorithms
Many many more….
29-08-2024 [email protected] 33
29-08-2024 [email protected] 34
29-08-2024 [email protected] 35
29-08-2024 [email protected] 36
29-08-2024 [email protected] 37
29-08-2024 [email protected] 38
29-08-2024 [email protected] 39
29-08-2024 [email protected] 40
The CIFAR-10 dataset (Canadian Institute For Advanced Research)
29-08-2024 [email protected] 41
What exactly is Deep Learning ?
• “Deep Learning” means using a neural network with several layers of
nodes between input and output
• The series of layers between input & output do feature identification
and processing in a series of stages, just as our brains seem to
• But,.. Multilayer neural networks have been around for 25 years.
• What’s actually new?
29-08-2024
[email protected] 42
29-08-2024
[email protected] 43
What exactly is Deep Learning ?
• We have always had good algorithms for learning the weights in networks with 1 hidden layer
• But these algorithms are not good at learning the weights for networks with more hidden
layers
• What’s new is: algorithms for training many-layer networks
29-08-2024 [email protected] 44
Machine Learning vs Deep Learning
29-08-2024 [email protected] 45
Motivation for Deep Learning
• Deep Architectures can be representationally efficient
• Fewer computational units for same function
• Deep Representations might allow for a hierarchy or representation
• Allows non-local generalization
• Comprehensibility
• Multiple levels of latent variables allow combinatorial sharing of
statistical strength
• Deep architectures work well (vision, audio, NLP, etc.)
29-08-2024 [email protected] 46
29-08-2024 [email protected] 47
Machine Learning vs Deep Learning
29-08-2024 [email protected] 48
Deep Learning Packages
29-08-2024 [email protected] 49
29-08-2024 [email protected] 50
29-08-2024 [email protected] 51
29-08-2024 [email protected] 52
29-08-2024 [email protected] 53
29-08-2024 [email protected] 54
29-08-2024 [email protected] 55
29-08-2024 [email protected] 56
29-08-2024 [email protected] 57
29-08-2024 [email protected] 58
Mimic of Five Senses
Vision
59
29-08-2024 [email protected]
If I asked you to name the objects in the picture
“tablecloth, basket, grass, boy, girl, man, woman, orange juice
bottle, tomatoes, lettuce, disposable plates…”
if I told you to describe the picture
“It’s the picture of a family picnic”
29-08-2024 [email protected] 60
Computer vision is a field of artificial intelligence that trains computers to
interpret and understand the visual world.
Using digital images from cameras and videos and deep learning models,
machines can accurately identify and classify objects — and then react to what
they “see.”
29-08-2024 [email protected] 61
What is Computer Vision(CV)
• Computer vision, or the ability of artificially intelligent systems to “see”
like humans, has been a subject of increasing interest and rigorous
research for decades now
• As a way of emulating the human visual system, the research in the field
of computer vision purports to develop machines that can automate
tasks that require visual cognition
• However, the process of deciphering images, due to the significantly
greater amount of multi-dimensional data that needs analysis, is much
more complex than understanding other forms of binary information
• This makes developing AI systems that can recognize visual data more
complicated
29-08-2024 [email protected] 62
Vision Based Sensors
Google Lens
Translation
Google is back with the Google googles — except it uses advanced photo
recognition technology, and it’s called Google Lens
29-08-2024 [email protected] 63
29-08-2024 [email protected] 64
FC_I5.0 [email protected] 65
Identify faces and expressions
29-08-2024 [email protected] 66
29-08-2024 [email protected] 67
29-08-2024 [email protected] 68
29-08-2024 [email protected] 69
29-08-2024 [email protected] 70
Everyday Examples of Artificial Intelligence and
Deep Learning Self Driving Cars
29-08-2024 [email protected] 71
29-08-2024 [email protected] 72
29-08-2024 [email protected] 73
29-08-2024 [email protected] 74
29-08-2024 [email protected] 75
29-08-2024 [email protected] 76
29-08-2024 [email protected] 77
29-08-2024 [email protected] 78
FACIAL EXPERSSION RECOGNITION
Architecture of proposed DCNN, R: Regularizer,
F: Number of filters, K: Kernel size
(a) Intra-class variations (b) Inter-class similarities
Proposed DCNN-VC Model
Effect of classification accuracy based on (a) learning rates, (b) mini-batch size, (c)
optimizers and (d) number of epochs
Overall accuracy of proposed methods on five datasets (%)
Publications
1. Ch. Venkata Rami Reddy , U. Srinivasulu Reddy, “Virtual Facial Expression Recognition using Deep CNN
with Ensemble Learning”, Journal of Ambient Intelligence and Humanized Computing, 12 (3), (2021):
10581-10599 DOI: 10.1007/s12652-020-02866-3, (SCIE, IF: 3.662, Q1 Journal).
2. Ch. Venkata Rami Reddy, U. Srinivasulu Reddy. “Facial Emotion Recognition using NLPCA and SVM”,
Traitement du Signal, 36 (1), (2019): 13-22. https://doi.org/10.18280/ts.360102, (SCIE, IF: 2.299, Q3
Journal).
79
29-08-2024 [email protected]
FACIAL EXPERSSION RECOGNITION
Compound facial expression samples Proposed frame work for CFER
Performance comparison of various classifiers Comparison of the proposed method with existing
Classifier Accuracy (%) methods
SVM 90.11 Author/Method Dataset Accuracy (%)
Random Forest 86.60
Du S et al. (2014) CFED 76.91
Gaussian NB 73.18
Logistic Regression 88.86 Slimani et al. (2019) CFED 52.00
KNN 84.65 Zhao Y et al. (2019) CMED 67.15
Decision tree 62.97 Proposed Model CFED 90.11
80
29-08-2024 [email protected]
Project: Development of Computer Vision System for an AI assistant - 2 years (2022-2024)
Funded By: Space Technology Incubation Centre, ISRO
29-08-2024 [email protected] 81
29-08-2024 [email protected] 82
Smart HEALTH CARE
29-08-2024 [email protected] 85
29-08-2024 [email protected] 86
29-08-2024 [email protected] 87
29-08-2024 [email protected] 88
29-08-2024 [email protected] 89
29-08-2024 [email protected] 90
29-08-2024 [email protected] 91
29-08-2024 [email protected] 92
29-08-2024 [email protected] 93
29-08-2024 [email protected] 94
29-08-2024 [email protected] 95
29-08-2024 [email protected] 96
29-08-2024 [email protected] 97
29-08-2024 [email protected] 98
29-08-2024 [email protected] 99
29-08-2024 [email protected] 100
CLASSIFICATION, DETECTION AND RECOGNITION OF INSECTS IN FIELD CROPS
101
29-08-2024 [email protected]
CLASSIFICATION, DETECTION AND RECOGNITION OF INSECTS IN FIELD CROPS
Supported By: DST, Govt. of India
Grant No.: DST/Disha/SoRF-PM/059/2013(G)
Publications
1. K. Thenmozhi, S. Dakshayani, U. Srinivasulu Reddy, “Insect classification and detection in
field crops using modern machine learning techniques’, Information Processing in
Agriculture, Elsevier. Article in Press, DOI: 10.1016/j.inpa.2020.09.006 (Scopus, IS: 7.42,
Q1 Journal)
2. K. Thenmozhi, U. Srinivasulu Reddy, “Machine learning ensemble with image processing
for pest identification and classification in field crops”, Neural Computing and Applications.
Springer, Accepted, (SCIE, IF: 5.102, Q1 Journal)
102
29-08-2024 [email protected]
CROP PEST CLASSIFICATION BASED ON DEEP CNN AND TRANSFER LEARNING
Proposed CNN Architecture
Overall classification accuracies
Data set Xie et Cheng et al. Cheng et al. Xia et Xie et
al. (2017a) (2017b) al. al.
(2015) (24 classes) (24 classes) (2018) (2018) Proposed methods
(24 (24 (40
classes) classes) classes)
MKL AlexNet VGG- ResNet- ResNet- VGG-19 MKL Proposed Pre-trained models with transfer learning
16 50 101 + RPN CNN
AlexNet ResNet-50 ResNet- VGG- VGG-19
model 101 16
Xie1 91.2 86.67 95.33 94.67 98.67 89.22 - 97.47 94.23 95.87 95.95 96.25 95.89
(24
classes)
Xie2 - - - - - - 89.3 95.97 92.25 92.96 93.99 93.67 93.03
(40
classes)
Publications
1. K. Thenmozhi, U. Srinivasulu Reddy, “Crop pest classification based on deep convolutional neural network and transfer learning”,
Computers and Electronics in Agriculture 164 (2019): 104906. Elsevier. (SCIE, IF: 6.757, Q1 Journal) 103
29-08-2024 [email protected]
AUTOMATIC ROBUST RECOGNITION AND DETECTION OF INSECTS IN PADDY, MAIZE AND SUGARCANE
CROPS USING YOLO MODELS
Detection results of YOLO models
Comparison of Detection Results
Patent No:202241035191 dated
June 20, 2022 (provisional patent filed)
Detection results on FCI Test Dataset
Detection Insect BFLOPS TP FP FN P R F1- Avg mAP Average
Method image size (%) (%) score IoU (%) detection
(pixels) (%) (%) Time (sec)
YOLOv2 416 ×416 34.916 341 91 63 78.9 84.4 81.5 78.44 91.67 0.0283
YOLOv3 416 ×416 65.457 496 57 32 89.6 93.9 91.7 89.38 96.31 0.0407
YOLOv3-Tiny 416 ×416 5.483 381 79 45 82.8 89.4 85.97 87.35 94.14 0.0397
104
29-08-2024 [email protected]
ARTIFICIAL INTELLIGENCE BASED APPROACH FOR FIELD CROP
INSECT DETECTION [ETI-12]
105
29-08-2024 [email protected]
FC_I5.0 [email protected] 106
FC_I5.0 [email protected] 107
29-08-2024 [email protected] 108
Thank You
29-08-2024 109
[email protected]
