SlideShare a Scribd company logo
GRADIENT BASED LEARNING
APPLIED TO DOCUMENT
RECOGNITION
Submitted by- Gyanendra Awasthi
(Roll no. -201315)
INSTRUCTOR- Prof. Nishchal K. Verma
Intoduction
 Authored by-
• Yann LeCun
• Leon Botteou
• Yoshua Bengio
• Patrick Haffiner
 Presented in the proceedings of the IEEE November 1998
 Yann LeCun and other co-authors introduced LeNet-5 architecture
through this paper
 Main message of the paper is building better pattern recognition system
relying more on automatic learning than hand–design heuristic
Introduction
 Traditional Approach
• Feature Extraction Module
• Trainable Classifier Module
 Limitations to this approach
• Feature extraction is hand crafted
• Classifier use low dimensional spaces in learning
techniques
 Enabling of high dimensional spaces due to
• Availability of low cost machines relying on ‘numerical
methods’
• Availability of large database
• Availability of powerful machine learning techniques
Introduction
 Why Gradient based Learning method
• Easier to minimize a smooth continuous function than
discrete function
• Minimal processing requirements
 Gradient Back- Propagation
• This with sigmoidal units can solve complicated learning
tasks if applied to multi layer neural network
Convolution Neural Network(CNN)
 CNN is standard form of neural network architecture associated
with image recognition
 CNN architectures are most popular deep learning framework
 Its applications are in marketing, healthcare, retail, automotive
 Characteristics of CNN architectures are-
• Local Receptive Fields
• Sub-sampling
• Weight Sharing
Fig. A typical architecture of CNN
LeNet-5 Convolution Neural Network
 LeNet-5 Architecture comprises of 7 layers(not counting the input
layer)
• 3 Convolutional Layers (Cx)
• 2 Subsampling Layers (Sx)
• 2 Fully Connected Layers (Fx)
where x denotes the layer’s index.
LeNet-5 Architecture
 First Layer (C1)
• Convolutional layer with 6 feature map of size 28*28
• Each unit in each feature map is connected to a 5*5 neighborhood in the input
• Contains 156 trainable parameters and 122,304 connections
 Second Layer (S2)
• A subsampling layer with 6 feature map of size 14*14
• Each unit in each feature map is connected to a 2*2 neighborhood in the
corresponding feature map in C1
• Contains 12 trainable parameters and 5880 connections
 Third Layer (C3)
• A Convolutional layer with 16 feature map of size 28*28
• Each unit in each feature map is connected to a 5*5 neighborhoods at identical
locations in S2 feature map
LeNet-5 Architecture
 Fourth Layer (S4)
• A subsampling layer with 16 feature map of size 5*5
• Each unit in each feature map is connected to a 2*2
neighborhood in the corresponding feature map in C3
• Contains 32 trainable parameters and 2000 connections
 Fifth Layer (C5)
• A convolutional layer with 120 feature map of size 1*1
• Each unit in each feature map is connected to a 5*5
neighborhood on all 16 of S4’s features
• Has 48120 trainable connections
 Sixth Layer (F6)
• A fully connected layer contains 84 units and fully connected to
C5
• Has 10164 trainable parameters
LeNet-5 Architecture
LeNet-1 Architecture
 Consists of
 3 convolutional layers
 2 Subsampling layers
 Number of parameters is about 3000
 The architecture is as such
 28×28 input image
 Four 24×24 feature maps
convolutional layer (5×5 size)
 Average Pooling layers (2×2 size)
 Eight 12×12 feature maps
convolutional layer (5×5 size)
 Average Pooling layers (2×2 size)
 Directly fully connected to the output
Fig. LeNet-1 Architecture
LeNet-4 Architecture
 Consists of :
 3 convolutional layers
 2 Subsampling layers
 1 Full connection layers
 Contains about 260,000 connections and 17,000 free parameters
 In LeNet-4, the input is 32*32 input layer in which 20*20 images (not
deslanted ) were centred by centre of mass.
Database- Modified NIST Dataset
 NIST- National Institute of Standard and Technology
database
 MNIST database
• Consists of handwritten images of digits from 0
to 9
• Subset of famous NIST Dataset
 Images centered in 28*28 pixels in black and white
 Dataset of 70,000 images of which 60,000 are for
training and remaining 10,000 for test set
MNIST Dataset
Training Set,
60000
Test set,
10000
Fig. MNIST Dataset
Some Training Images
Results: LeNet-1
• MNIST database is used in which
10% dataset is used for validation
and remaining for training.
• Test loss: 0.05995325744152069
• Test accuracy:
0.9811999797821045
• As number of epochs increases,
training loss and validation loss
decreases. Also training accuracy
and validation accuracy increases
with number of epochs.
Results: LeNet-4
• Test loss: 0.052783720195293427
• Test accuracy:
0.9832000136375427
• As number of epochs increases,
training loss and validation loss
decreases. Also training accuracy
and validation accuracy increases
with number of epochs.
• The test accuracy of LeNet-4 is
more than LeNet-1. Also, test loss
is less than of LeNet-1.
Results: LeNet-5
• Test loss: 0.038686320185661316
• Test accuracy:
0.9866999983787537
• As number of epochs increases,
training loss and validation loss
decreases. Also training accuracy
and validation accuracy increases
with number of epochs.
• The test accuracy of LeNet-5 is
more than both of LeNet-4 and
LeNet-1. Also, test loss is lesser
than both of LeNet-1and LeNet-4.
Results:
Comparison of various classifiers on MNIST
Dataset
Gradient Based Learning Applied to Document Recognition

More Related Content

What's hot

4. social network analysis
4. social network analysis4. social network analysis
4. social network analysis
Lokesh Ramaswamy
 
Image classification using CNN
Image classification using CNNImage classification using CNN
Image classification using CNN
Noura Hussein
 
Social Recommender Systems
Social Recommender SystemsSocial Recommender Systems
Social Recommender Systems
guest77b0cd12
 
Ensemble methods
Ensemble methodsEnsemble methods
Ensemble methods
Christopher Marker
 
Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...
Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...
Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...
Sujit Pal
 
Introduction to Neural Networks
Introduction to Neural NetworksIntroduction to Neural Networks
Introduction to Neural Networks
Databricks
 
Regularization in deep learning
Regularization in deep learningRegularization in deep learning
Regularization in deep learning
Kien Le
 
Curse of Dimensionality and Big Data
Curse of Dimensionality and Big DataCurse of Dimensionality and Big Data
Curse of Dimensionality and Big Data
Stephane Marchand-Maillet
 
Lecture 4 Decision Trees (2): Entropy, Information Gain, Gain Ratio
Lecture 4 Decision Trees (2): Entropy, Information Gain, Gain RatioLecture 4 Decision Trees (2): Entropy, Information Gain, Gain Ratio
Lecture 4 Decision Trees (2): Entropy, Information Gain, Gain Ratio
Marina Santini
 
House price prediction
House price predictionHouse price prediction
House price prediction
AdityaKumar1505
 
An Introduction To Weka
An Introduction To WekaAn Introduction To Weka
An Introduction To Weka
weka Content
 
Training Neural Networks
Training Neural NetworksTraining Neural Networks
Training Neural Networks
Databricks
 
Master's Thesis Presentation
Master's Thesis PresentationMaster's Thesis Presentation
Master's Thesis Presentation
Wajdi Khattel
 
Classification by back propagation, multi layered feed forward neural network...
Classification by back propagation, multi layered feed forward neural network...Classification by back propagation, multi layered feed forward neural network...
Classification by back propagation, multi layered feed forward neural network...
bihira aggrey
 
Artificial neural network
Artificial neural networkArtificial neural network
Artificial neural network
mustafa aadel
 
Deep learning - A Visual Introduction
Deep learning - A Visual IntroductionDeep learning - A Visual Introduction
Deep learning - A Visual Introduction
Lukas Masuch
 
“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...
“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...
“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...
Edge AI and Vision Alliance
 
Architecture Design for Deep Neural Networks III
Architecture Design for Deep Neural Networks IIIArchitecture Design for Deep Neural Networks III
Architecture Design for Deep Neural Networks III
Wanjin Yu
 
Image classification using cnn
Image classification using cnnImage classification using cnn
Image classification using cnn
Debarko De
 
Introduction to Deep learning
Introduction to Deep learningIntroduction to Deep learning
Introduction to Deep learning
leopauly
 

What's hot (20)

4. social network analysis
4. social network analysis4. social network analysis
4. social network analysis
 
Image classification using CNN
Image classification using CNNImage classification using CNN
Image classification using CNN
 
Social Recommender Systems
Social Recommender SystemsSocial Recommender Systems
Social Recommender Systems
 
Ensemble methods
Ensemble methodsEnsemble methods
Ensemble methods
 
Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...
Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...
Transfer Learning and Fine Tuning for Cross Domain Image Classification with ...
 
Introduction to Neural Networks
Introduction to Neural NetworksIntroduction to Neural Networks
Introduction to Neural Networks
 
Regularization in deep learning
Regularization in deep learningRegularization in deep learning
Regularization in deep learning
 
Curse of Dimensionality and Big Data
Curse of Dimensionality and Big DataCurse of Dimensionality and Big Data
Curse of Dimensionality and Big Data
 
Lecture 4 Decision Trees (2): Entropy, Information Gain, Gain Ratio
Lecture 4 Decision Trees (2): Entropy, Information Gain, Gain RatioLecture 4 Decision Trees (2): Entropy, Information Gain, Gain Ratio
Lecture 4 Decision Trees (2): Entropy, Information Gain, Gain Ratio
 
House price prediction
House price predictionHouse price prediction
House price prediction
 
An Introduction To Weka
An Introduction To WekaAn Introduction To Weka
An Introduction To Weka
 
Training Neural Networks
Training Neural NetworksTraining Neural Networks
Training Neural Networks
 
Master's Thesis Presentation
Master's Thesis PresentationMaster's Thesis Presentation
Master's Thesis Presentation
 
Classification by back propagation, multi layered feed forward neural network...
Classification by back propagation, multi layered feed forward neural network...Classification by back propagation, multi layered feed forward neural network...
Classification by back propagation, multi layered feed forward neural network...
 
Artificial neural network
Artificial neural networkArtificial neural network
Artificial neural network
 
Deep learning - A Visual Introduction
Deep learning - A Visual IntroductionDeep learning - A Visual Introduction
Deep learning - A Visual Introduction
 
“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...
“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...
“An Introduction to Data Augmentation Techniques in ML Frameworks,” a Present...
 
Architecture Design for Deep Neural Networks III
Architecture Design for Deep Neural Networks IIIArchitecture Design for Deep Neural Networks III
Architecture Design for Deep Neural Networks III
 
Image classification using cnn
Image classification using cnnImage classification using cnn
Image classification using cnn
 
Introduction to Deep learning
Introduction to Deep learningIntroduction to Deep learning
Introduction to Deep learning
 

Similar to Gradient Based Learning Applied to Document Recognition

Modern Convolutional Neural Network techniques for image segmentation
Modern Convolutional Neural Network techniques for image segmentationModern Convolutional Neural Network techniques for image segmentation
Modern Convolutional Neural Network techniques for image segmentation
Gioele Ciaparrone
 
Convolutional Neural Networks
Convolutional Neural NetworksConvolutional Neural Networks
Convolutional Neural Networks
milad abbasi
 
Cnn
CnnCnn
Internetworking fundamentals(networking)
Internetworking fundamentals(networking)Internetworking fundamentals(networking)
Internetworking fundamentals(networking)
welcometofacebook
 
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr..."Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
Edge AI and Vision Alliance
 
Chapter10.pptx
Chapter10.pptxChapter10.pptx
Chapter10.pptx
adnansbp
 
Unsupervised learning networks
Unsupervised learning networksUnsupervised learning networks
Unsupervised learning networks
Dr. C.V. Suresh Babu
 
Deep Learning Initiative @ NECSTLab
Deep Learning Initiative @ NECSTLabDeep Learning Initiative @ NECSTLab
Deep Learning Initiative @ NECSTLab
NECST Lab @ Politecnico di Milano
 
[2020 CVPR Efficient DET paper review]
[2020 CVPR Efficient DET paper review][2020 CVPR Efficient DET paper review]
[2020 CVPR Efficient DET paper review]
taeseon ryu
 
Convolutional Neural Networks : Popular Architectures
Convolutional Neural Networks : Popular ArchitecturesConvolutional Neural Networks : Popular Architectures
Convolutional Neural Networks : Popular Architectures
ananth
 
IRJET - Image Classification using CNN
IRJET - Image Classification using CNNIRJET - Image Classification using CNN
IRJET - Image Classification using CNN
IRJET Journal
 
"Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio...
"Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio..."Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio...
"Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio...
Edge AI and Vision Alliance
 
Deep_Learning_Frameworks_CNTK_PyTorch
Deep_Learning_Frameworks_CNTK_PyTorchDeep_Learning_Frameworks_CNTK_PyTorch
Deep_Learning_Frameworks_CNTK_PyTorch
Subhashis Hazarika
 
3_Transfer_Learning.pdf
3_Transfer_Learning.pdf3_Transfer_Learning.pdf
3_Transfer_Learning.pdf
FEG
 
NVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digits
NVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digitsNVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digits
NVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digits
NVIDIA Taiwan
 
Introduction to computer vision with Convoluted Neural Networks
Introduction to computer vision with Convoluted Neural NetworksIntroduction to computer vision with Convoluted Neural Networks
Introduction to computer vision with Convoluted Neural Networks
MarcinJedyk
 
Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...
Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...
Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...
changedaeoh
 
Introduction to computer vision
Introduction to computer visionIntroduction to computer vision
Introduction to computer vision
Marcin Jedyk
 
U-Netpresentation.pptx
U-Netpresentation.pptxU-Netpresentation.pptx
U-Netpresentation.pptx
NoorUlHaq47
 
intro-to-cnn-April_2020.pptx
intro-to-cnn-April_2020.pptxintro-to-cnn-April_2020.pptx
intro-to-cnn-April_2020.pptx
ssuser3aa461
 

Similar to Gradient Based Learning Applied to Document Recognition (20)

Modern Convolutional Neural Network techniques for image segmentation
Modern Convolutional Neural Network techniques for image segmentationModern Convolutional Neural Network techniques for image segmentation
Modern Convolutional Neural Network techniques for image segmentation
 
Convolutional Neural Networks
Convolutional Neural NetworksConvolutional Neural Networks
Convolutional Neural Networks
 
Cnn
CnnCnn
Cnn
 
Internetworking fundamentals(networking)
Internetworking fundamentals(networking)Internetworking fundamentals(networking)
Internetworking fundamentals(networking)
 
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr..."Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
 
Chapter10.pptx
Chapter10.pptxChapter10.pptx
Chapter10.pptx
 
Unsupervised learning networks
Unsupervised learning networksUnsupervised learning networks
Unsupervised learning networks
 
Deep Learning Initiative @ NECSTLab
Deep Learning Initiative @ NECSTLabDeep Learning Initiative @ NECSTLab
Deep Learning Initiative @ NECSTLab
 
[2020 CVPR Efficient DET paper review]
[2020 CVPR Efficient DET paper review][2020 CVPR Efficient DET paper review]
[2020 CVPR Efficient DET paper review]
 
Convolutional Neural Networks : Popular Architectures
Convolutional Neural Networks : Popular ArchitecturesConvolutional Neural Networks : Popular Architectures
Convolutional Neural Networks : Popular Architectures
 
IRJET - Image Classification using CNN
IRJET - Image Classification using CNNIRJET - Image Classification using CNN
IRJET - Image Classification using CNN
 
"Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio...
"Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio..."Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio...
"Quantizing Deep Networks for Efficient Inference at the Edge," a Presentatio...
 
Deep_Learning_Frameworks_CNTK_PyTorch
Deep_Learning_Frameworks_CNTK_PyTorchDeep_Learning_Frameworks_CNTK_PyTorch
Deep_Learning_Frameworks_CNTK_PyTorch
 
3_Transfer_Learning.pdf
3_Transfer_Learning.pdf3_Transfer_Learning.pdf
3_Transfer_Learning.pdf
 
NVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digits
NVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digitsNVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digits
NVIDIA 深度學習教育機構 (DLI): Medical image segmentation using digits
 
Introduction to computer vision with Convoluted Neural Networks
Introduction to computer vision with Convoluted Neural NetworksIntroduction to computer vision with Convoluted Neural Networks
Introduction to computer vision with Convoluted Neural Networks
 
Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...
Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...
Vision Transformer(ViT) / An Image is Worth 16*16 Words: Transformers for Ima...
 
Introduction to computer vision
Introduction to computer visionIntroduction to computer vision
Introduction to computer vision
 
U-Netpresentation.pptx
U-Netpresentation.pptxU-Netpresentation.pptx
U-Netpresentation.pptx
 
intro-to-cnn-April_2020.pptx
intro-to-cnn-April_2020.pptxintro-to-cnn-April_2020.pptx
intro-to-cnn-April_2020.pptx
 

More from Gyanendra Awasthi

Modelling Frontier Mortality using Bayesian Generalised Additive Models
Modelling Frontier Mortality using Bayesian Generalised Additive ModelsModelling Frontier Mortality using Bayesian Generalised Additive Models
Modelling Frontier Mortality using Bayesian Generalised Additive Models
Gyanendra Awasthi
 
Image Compression using K-Means Clustering Method
Image Compression using K-Means Clustering MethodImage Compression using K-Means Clustering Method
Image Compression using K-Means Clustering Method
Gyanendra Awasthi
 
Hyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolutionHyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolution
Gyanendra Awasthi
 
Study and performance analysis of combustion chamber using
Study and performance analysis of combustion chamber usingStudy and performance analysis of combustion chamber using
Study and performance analysis of combustion chamber using
Gyanendra Awasthi
 
Study and performance analysis of combustion chamber using ANSYS
Study and performance analysis of combustion chamber using ANSYSStudy and performance analysis of combustion chamber using ANSYS
Study and performance analysis of combustion chamber using ANSYS
Gyanendra Awasthi
 
Manufacturing of Stuffing box
Manufacturing of Stuffing boxManufacturing of Stuffing box
Manufacturing of Stuffing box
Gyanendra Awasthi
 
Microstructure and chemical compositions of ferritic stainless steel
Microstructure and chemical compositions of ferritic stainless steelMicrostructure and chemical compositions of ferritic stainless steel
Microstructure and chemical compositions of ferritic stainless steel
Gyanendra Awasthi
 
Hyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolutionHyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolution
Gyanendra Awasthi
 

More from Gyanendra Awasthi (8)

Modelling Frontier Mortality using Bayesian Generalised Additive Models
Modelling Frontier Mortality using Bayesian Generalised Additive ModelsModelling Frontier Mortality using Bayesian Generalised Additive Models
Modelling Frontier Mortality using Bayesian Generalised Additive Models
 
Image Compression using K-Means Clustering Method
Image Compression using K-Means Clustering MethodImage Compression using K-Means Clustering Method
Image Compression using K-Means Clustering Method
 
Hyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolutionHyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolution
 
Study and performance analysis of combustion chamber using
Study and performance analysis of combustion chamber usingStudy and performance analysis of combustion chamber using
Study and performance analysis of combustion chamber using
 
Study and performance analysis of combustion chamber using ANSYS
Study and performance analysis of combustion chamber using ANSYSStudy and performance analysis of combustion chamber using ANSYS
Study and performance analysis of combustion chamber using ANSYS
 
Manufacturing of Stuffing box
Manufacturing of Stuffing boxManufacturing of Stuffing box
Manufacturing of Stuffing box
 
Microstructure and chemical compositions of ferritic stainless steel
Microstructure and chemical compositions of ferritic stainless steelMicrostructure and chemical compositions of ferritic stainless steel
Microstructure and chemical compositions of ferritic stainless steel
 
Hyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolutionHyperloop- A 21st century transportation revolution
Hyperloop- A 21st century transportation revolution
 

Recently uploaded

The Ipsos - AI - Monitor 2024 Report.pdf
The  Ipsos - AI - Monitor 2024 Report.pdfThe  Ipsos - AI - Monitor 2024 Report.pdf
The Ipsos - AI - Monitor 2024 Report.pdf
Social Samosa
 
Udemy_2024_Global_Learning_Skills_Trends_Report (1).pdf
Udemy_2024_Global_Learning_Skills_Trends_Report (1).pdfUdemy_2024_Global_Learning_Skills_Trends_Report (1).pdf
Udemy_2024_Global_Learning_Skills_Trends_Report (1).pdf
Fernanda Palhano
 
一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理
一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理
一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理
nyfuhyz
 
Analysis insight about a Flyball dog competition team's performance
Analysis insight about a Flyball dog competition team's performanceAnalysis insight about a Flyball dog competition team's performance
Analysis insight about a Flyball dog competition team's performance
roli9797
 
4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...
4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...
4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...
Social Samosa
 
Open Source Contributions to Postgres: The Basics POSETTE 2024
Open Source Contributions to Postgres: The Basics POSETTE 2024Open Source Contributions to Postgres: The Basics POSETTE 2024
Open Source Contributions to Postgres: The Basics POSETTE 2024
ElizabethGarrettChri
 
一比一原版(UO毕业证)渥太华大学毕业证如何办理
一比一原版(UO毕业证)渥太华大学毕业证如何办理一比一原版(UO毕业证)渥太华大学毕业证如何办理
一比一原版(UO毕业证)渥太华大学毕业证如何办理
aqzctr7x
 
一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理
一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理
一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理
nuttdpt
 
办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样
办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样
办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样
apvysm8
 
A presentation that explain the Power BI Licensing
A presentation that explain the Power BI LicensingA presentation that explain the Power BI Licensing
A presentation that explain the Power BI Licensing
AlessioFois2
 
在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样
在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样
在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样
v7oacc3l
 
ViewShift: Hassle-free Dynamic Policy Enforcement for Every Data Lake
ViewShift: Hassle-free Dynamic Policy Enforcement for Every Data LakeViewShift: Hassle-free Dynamic Policy Enforcement for Every Data Lake
ViewShift: Hassle-free Dynamic Policy Enforcement for Every Data Lake
Walaa Eldin Moustafa
 
Everything you wanted to know about LIHTC
Everything you wanted to know about LIHTCEverything you wanted to know about LIHTC
Everything you wanted to know about LIHTC
Roger Valdez
 
原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理
原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理
原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理
wyddcwye1
 
06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM
06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM
06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM
Timothy Spann
 
DSSML24_tspann_CodelessGenerativeAIPipelines
DSSML24_tspann_CodelessGenerativeAIPipelinesDSSML24_tspann_CodelessGenerativeAIPipelines
DSSML24_tspann_CodelessGenerativeAIPipelines
Timothy Spann
 
Learn SQL from basic queries to Advance queries
Learn SQL from basic queries to Advance queriesLearn SQL from basic queries to Advance queries
Learn SQL from basic queries to Advance queries
manishkhaire30
 
原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样
原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样
原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样
ihavuls
 
一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理
一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理
一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理
xclpvhuk
 
Palo Alto Cortex XDR presentation .......
Palo Alto Cortex XDR presentation .......Palo Alto Cortex XDR presentation .......
Palo Alto Cortex XDR presentation .......
Sachin Paul
 

Recently uploaded (20)

The Ipsos - AI - Monitor 2024 Report.pdf
The  Ipsos - AI - Monitor 2024 Report.pdfThe  Ipsos - AI - Monitor 2024 Report.pdf
The Ipsos - AI - Monitor 2024 Report.pdf
 
Udemy_2024_Global_Learning_Skills_Trends_Report (1).pdf
Udemy_2024_Global_Learning_Skills_Trends_Report (1).pdfUdemy_2024_Global_Learning_Skills_Trends_Report (1).pdf
Udemy_2024_Global_Learning_Skills_Trends_Report (1).pdf
 
一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理
一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理
一比一原版(UMN文凭证书)明尼苏达大学毕业证如何办理
 
Analysis insight about a Flyball dog competition team's performance
Analysis insight about a Flyball dog competition team's performanceAnalysis insight about a Flyball dog competition team's performance
Analysis insight about a Flyball dog competition team's performance
 
4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...
4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...
4th Modern Marketing Reckoner by MMA Global India & Group M: 60+ experts on W...
 
Open Source Contributions to Postgres: The Basics POSETTE 2024
Open Source Contributions to Postgres: The Basics POSETTE 2024Open Source Contributions to Postgres: The Basics POSETTE 2024
Open Source Contributions to Postgres: The Basics POSETTE 2024
 
一比一原版(UO毕业证)渥太华大学毕业证如何办理
一比一原版(UO毕业证)渥太华大学毕业证如何办理一比一原版(UO毕业证)渥太华大学毕业证如何办理
一比一原版(UO毕业证)渥太华大学毕业证如何办理
 
一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理
一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理
一比一原版(UCSF文凭证书)旧金山分校毕业证如何办理
 
办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样
办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样
办(uts毕业证书)悉尼科技大学毕业证学历证书原版一模一样
 
A presentation that explain the Power BI Licensing
A presentation that explain the Power BI LicensingA presentation that explain the Power BI Licensing
A presentation that explain the Power BI Licensing
 
在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样
在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样
在线办理(英国UCA毕业证书)创意艺术大学毕业证在读证明一模一样
 
ViewShift: Hassle-free Dynamic Policy Enforcement for Every Data Lake
ViewShift: Hassle-free Dynamic Policy Enforcement for Every Data LakeViewShift: Hassle-free Dynamic Policy Enforcement for Every Data Lake
ViewShift: Hassle-free Dynamic Policy Enforcement for Every Data Lake
 
Everything you wanted to know about LIHTC
Everything you wanted to know about LIHTCEverything you wanted to know about LIHTC
Everything you wanted to know about LIHTC
 
原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理
原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理
原版一比一利兹贝克特大学毕业证(LeedsBeckett毕业证书)如何办理
 
06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM
06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM
06-12-2024-BudapestDataForum-BuildingReal-timePipelineswithFLaNK AIM
 
DSSML24_tspann_CodelessGenerativeAIPipelines
DSSML24_tspann_CodelessGenerativeAIPipelinesDSSML24_tspann_CodelessGenerativeAIPipelines
DSSML24_tspann_CodelessGenerativeAIPipelines
 
Learn SQL from basic queries to Advance queries
Learn SQL from basic queries to Advance queriesLearn SQL from basic queries to Advance queries
Learn SQL from basic queries to Advance queries
 
原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样
原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样
原版制作(unimelb毕业证书)墨尔本大学毕业证Offer一模一样
 
一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理
一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理
一比一原版(Unimelb毕业证书)墨尔本大学毕业证如何办理
 
Palo Alto Cortex XDR presentation .......
Palo Alto Cortex XDR presentation .......Palo Alto Cortex XDR presentation .......
Palo Alto Cortex XDR presentation .......
 

Gradient Based Learning Applied to Document Recognition

  • 1. GRADIENT BASED LEARNING APPLIED TO DOCUMENT RECOGNITION Submitted by- Gyanendra Awasthi (Roll no. -201315) INSTRUCTOR- Prof. Nishchal K. Verma
  • 2. Intoduction  Authored by- • Yann LeCun • Leon Botteou • Yoshua Bengio • Patrick Haffiner  Presented in the proceedings of the IEEE November 1998  Yann LeCun and other co-authors introduced LeNet-5 architecture through this paper  Main message of the paper is building better pattern recognition system relying more on automatic learning than hand–design heuristic
  • 3. Introduction  Traditional Approach • Feature Extraction Module • Trainable Classifier Module  Limitations to this approach • Feature extraction is hand crafted • Classifier use low dimensional spaces in learning techniques  Enabling of high dimensional spaces due to • Availability of low cost machines relying on ‘numerical methods’ • Availability of large database • Availability of powerful machine learning techniques
  • 4. Introduction  Why Gradient based Learning method • Easier to minimize a smooth continuous function than discrete function • Minimal processing requirements  Gradient Back- Propagation • This with sigmoidal units can solve complicated learning tasks if applied to multi layer neural network
  • 5. Convolution Neural Network(CNN)  CNN is standard form of neural network architecture associated with image recognition  CNN architectures are most popular deep learning framework  Its applications are in marketing, healthcare, retail, automotive  Characteristics of CNN architectures are- • Local Receptive Fields • Sub-sampling • Weight Sharing Fig. A typical architecture of CNN
  • 6. LeNet-5 Convolution Neural Network  LeNet-5 Architecture comprises of 7 layers(not counting the input layer) • 3 Convolutional Layers (Cx) • 2 Subsampling Layers (Sx) • 2 Fully Connected Layers (Fx) where x denotes the layer’s index.
  • 7. LeNet-5 Architecture  First Layer (C1) • Convolutional layer with 6 feature map of size 28*28 • Each unit in each feature map is connected to a 5*5 neighborhood in the input • Contains 156 trainable parameters and 122,304 connections  Second Layer (S2) • A subsampling layer with 6 feature map of size 14*14 • Each unit in each feature map is connected to a 2*2 neighborhood in the corresponding feature map in C1 • Contains 12 trainable parameters and 5880 connections  Third Layer (C3) • A Convolutional layer with 16 feature map of size 28*28 • Each unit in each feature map is connected to a 5*5 neighborhoods at identical locations in S2 feature map
  • 8. LeNet-5 Architecture  Fourth Layer (S4) • A subsampling layer with 16 feature map of size 5*5 • Each unit in each feature map is connected to a 2*2 neighborhood in the corresponding feature map in C3 • Contains 32 trainable parameters and 2000 connections  Fifth Layer (C5) • A convolutional layer with 120 feature map of size 1*1 • Each unit in each feature map is connected to a 5*5 neighborhood on all 16 of S4’s features • Has 48120 trainable connections  Sixth Layer (F6) • A fully connected layer contains 84 units and fully connected to C5 • Has 10164 trainable parameters
  • 10. LeNet-1 Architecture  Consists of  3 convolutional layers  2 Subsampling layers  Number of parameters is about 3000  The architecture is as such  28×28 input image  Four 24×24 feature maps convolutional layer (5×5 size)  Average Pooling layers (2×2 size)  Eight 12×12 feature maps convolutional layer (5×5 size)  Average Pooling layers (2×2 size)  Directly fully connected to the output Fig. LeNet-1 Architecture
  • 11. LeNet-4 Architecture  Consists of :  3 convolutional layers  2 Subsampling layers  1 Full connection layers  Contains about 260,000 connections and 17,000 free parameters  In LeNet-4, the input is 32*32 input layer in which 20*20 images (not deslanted ) were centred by centre of mass.
  • 12. Database- Modified NIST Dataset  NIST- National Institute of Standard and Technology database  MNIST database • Consists of handwritten images of digits from 0 to 9 • Subset of famous NIST Dataset  Images centered in 28*28 pixels in black and white  Dataset of 70,000 images of which 60,000 are for training and remaining 10,000 for test set MNIST Dataset Training Set, 60000 Test set, 10000 Fig. MNIST Dataset
  • 14. Results: LeNet-1 • MNIST database is used in which 10% dataset is used for validation and remaining for training. • Test loss: 0.05995325744152069 • Test accuracy: 0.9811999797821045 • As number of epochs increases, training loss and validation loss decreases. Also training accuracy and validation accuracy increases with number of epochs.
  • 15. Results: LeNet-4 • Test loss: 0.052783720195293427 • Test accuracy: 0.9832000136375427 • As number of epochs increases, training loss and validation loss decreases. Also training accuracy and validation accuracy increases with number of epochs. • The test accuracy of LeNet-4 is more than LeNet-1. Also, test loss is less than of LeNet-1.
  • 16. Results: LeNet-5 • Test loss: 0.038686320185661316 • Test accuracy: 0.9866999983787537 • As number of epochs increases, training loss and validation loss decreases. Also training accuracy and validation accuracy increases with number of epochs. • The test accuracy of LeNet-5 is more than both of LeNet-4 and LeNet-1. Also, test loss is lesser than both of LeNet-1and LeNet-4.
  • 17. Results: Comparison of various classifiers on MNIST Dataset