convolutional neural network and its application.pdf

Convolutional Neural network and its applications

1. ” ConvolutionalNeural Networks (CNN)”
University of Manouba.
Tunis Higher School of Business
Prepared by:
Sirine BEN AMMAR
2023-2024
Sirine BEN AMMAR

2. Outline
1General context
5CNN Components
Sirine BEN AMMAR 2/21
CNN Architecture
4
Applications of CNN models6
Implementation7
3Properties of CNN models
Definition
2
Some recent articles8
Conclusions9

3. Context
2
3
3/21Sirine BEN AMMAR

4. Context
2
3
4/21Sirine BEN AMMAR
Machine Learning VSDeep Learning

5. Context
2
3
5/21Sirine BEN AMMAR

6. Context
2
3
6/21Sirine BEN AMMAR

7. Definition
2
3
7/22Sirine BEN AMMAR
❑Convolutional Neural
Networks(CNNs)learns
multi-levelfeaturesand
classifierinajointfashionand
performsmuchbetterthan
traditionalapproachesfor
variousimageclassification
andsegmentationproblems.

8. Properties of CNN models
2
3
8/22Sirine BEN AMMAR
➢Sparse interactions between NN units (through kernels of small
size)
✓fewer parameters to learn
✓less computation resources are required
➢Parameter sharing (same kernel is applied throughout the input)
✓Maintain the same feature detection throughout the
input.
➢Ability to (automatically) learn local structure
➢Can handle variable-sized inputs.

9. CNN Architecture
2
3
9/22Sirine BEN AMMAR
➢Typically, a CNN model consists of convolution layers, for feature selection,
followed by fully connected layers that perform the prediction task.

10. CNN Architecture
2
3
10/22Sirine BEN AMMAR

11. CNN Components
2
3
11/22Sirine BEN AMMAR
Convolution
Non Linearity
Pooling or Sub Sampling
Classification (Fully Connected Layer)
➢There are 4 components in the CNN:

12. CNN Components

2
3
12/22Sirine BEN AMMAR
Input:
•Animageisamatrixofpixelvalues.
➢Ifweconsideragrayscaleimage,the
valueofeachpixelinthematrixwillrange
from0to255.
➢IfweconsideranRGBimage,
eachpixelwillhavethecombined
valuesofR,GandB.

13. CNN Components

2
3
13/22Sirine BEN AMMAR
Convolution Non Linearity Pooling Classification
1.Convolution :
➢TheprimarypurposeofconvolutionincaseofaCNNistoextract
featuresfromtheinputimage.

14. CNN Components
2
3
14/22Sirine BEN AMMAR
Convolution Non Linearity Pooling Classification
2.Non Linearity (ReLU):
➢ Replaces all negative pixel values in the
feature map by zero.
➢ The purpose of ReLUis to introduce
non-linearity in CNN, since most of the
real-world data would be non-linear.
➢ Other non-linear functions such as
tanh(-1, 1) or sigmoid (0, 1) can also
be used instead of ReLU(0, input).

15. CNN Components
2
3
15/22Sirine BEN AMMAR
Convolution Non Linearity Pooling Classification
3.Pooling:
➢ Reduces the dimensionality of each feature map but retains the most
important information.

16. CNN Components
3
12/22Sirine BEN AMMAR 16/ 22Sirine BEN AMMAR
Convolution Non Linearity Pooling Classification
4.Fully Connected Layer:
➢ The term “Fully Connected” implies that every neuron in the previous
layer is connected to every neuron on the next layer.
➢Theiractivationscanhencebecomputed
withamatrixmultiplicationfollowedbya
biasoffset.
➢Thepurposeofthefullyconnectedlayer
istousethehigh-levelfeaturesforclassifying
theinputimageintovariousclassesbasedon
thetrainingdataset.

17. Applications of CNN models
2
3
17/22Sirine BEN AMMAR
➢Image Processing
✓image classification
✓object detection
✓image segmentation
✓object tracking
✓face recognition…
➢Speech Processing
➢Text Detections and Recognition (OCR)
➢Natural Language Processing
➢Drug Discovery
➢TimeseriesAnalysis
✓Health risk assessment
✓Electromyography (EMG)
recognition…

18. Implementation
2
3
18/22Sirine BEN AMMAR
Link: https://colab.research.google.com/drive/17Svx0pQE_0g-
4uz22W_F0nFdsySiurt0

19. Some recent articles
2
3
19/22Sirine BEN AMMAR

20. Some recent articles
2
3
20/22Sirine BEN AMMAR

21. Conclusions
2
3
21/22
➢Inconclusion,ConvolutionalNeuralNetworksrepresentamajor
breakthroughindeeplearningwithvastandvariedapplications.
➢Aswewrapupthispresentation,let'slooktothefuture:new
challenges,technologicaladvancements,andextendedapplications.
➢Ongoingcommitmenttoresearchanddevelopmentiscrucialto
fullyharnessthepotentialofthisever-evolvingtechnology.
Sirine BEN AMMAR

22. 2
3
Sirine BEN AMMAR
Thank you!