SlideShare a Scribd company logo
1 of 9
SPOT THE MASK
CHALLENGE
INTRODUCTION
• This challenge involves employing Computer Vision to
identify whether a person is wearing nose masks or not.
• This challenge is to help ensure that masks are being
worn appropriately in public.
• The goal is to use data to create an image classification
model to accurately predict whether an image contains
a person wearing nose mask or not
20XX presentation title 2
MODEL PIPELINE
• DATA STRUCTURE ANALYSIS
• DATA PREPROCESSING
• DATA PREPARATION
• MODEL SELECTION
• MODEL TRAINING
• MODEL AND HYPERPARAMTER FINE-TUNING
• MODEL EVALUATION
20XX presentation title 3
METHODOLOGY(1)
• DATA STRUCTURE ANALYSIS
• USE OF PANDA DATAFRAME FUNCTIONS TO ANALYZE THE DATA FROM THE
TRAIN.CSV
• TRAIN AND TEST IMAGES BOTH GROUPED INTO ONE FOLDER
• DATA PREPROCESSING
• TRAIN AND TEST IMAGES BOTH GROUPED INTO ONE FOLDER HAD TO BE
SEPARATE FOLDERS
• USING THE TARGET IN THE TRAIN..CSV FILE, THE IMAGES WERE
SEPARATED INTO WITH NOSE MASKS AND WITHOUT NOSE MASKS
• DATA PREPARATION
• PREPROCESS THE DATA BY RESIZING THE IMAGES AND CREATED
DATAFRAMES FOR BOTH WITH NOSE MASKS AND WITHOUT NOSE MASKS
• USE AN IMAGE DATA GENERATOR FROM TENSORFLOW TO AUGMENT
THE DATA
• SPLIT THE DATASET AFTER AUGMENTATION INTO TWO USING
TRAIN_TEST_SPLIT
20XX presentation title 4
METHODOLOGY(2)
• MODEL SELECTION
• PRE-TRAINED MODELS SUCH AS VGG16, RESNET AND DENSET WERE
SELECTED FOR THEIR IMAGE CLASSIFICATION ABILITIES
• A CONVOLUTIONAL NEURAL NETWORK (CNN) IS BUILT USING THE PRE
TRAINED MODELS
• MODEL TRAINING
• USING THE PRE-TRAINED MODEL (DENSENET200), A CNN OF 3 LAYERS
IS BUILT WITH GLOBALAVERAGEPOOLING2D, DENSE AND DROPOUT
• THE MODEL IS TRAINED USING THE DATASET WITH ONE SET TRAIINING
AND THE OTHER SET EVALUATING IT
• MODEL AND HYPERPARAMETERS FINE-TUNING
• THE MODEL IS FINE-TUNED BY FREEZING SOME LAYERS AND TRAINING
THE REMAINING LAYERS
• HYPERPARAMETERS ARE OPTIMIZED AS WELL TO IMPROVE
PERFORMANCE
20XX presentation title 5
METHODOLOGY(3)
• MODEL EVALUATION
• TEST MODEL ON THE TEST DATASET PROVIDED TO ENSURE MODEL IS
WORKING
20XX presentation title 6
RESULTS
• THE CNN MODEL YIELED AN ACCURACY OF 0.95875 DURING
TRAINING AND IMPROVED TO 0.98769
• THE ACCURACY OF THE TEST WAS 0.990180639
20XX presentation title 7
RANKINGS
20XX presentation title 8
THANK YOU

More Related Content

Similar to SPOT THE MASK.pptx

Moving object detection in complex scene
Moving object detection in complex sceneMoving object detection in complex scene
Moving object detection in complex sceneKumar Mayank
 
Handwritten Digit Recognition and performance of various modelsation[autosaved]
Handwritten Digit Recognition and performance of various modelsation[autosaved]Handwritten Digit Recognition and performance of various modelsation[autosaved]
Handwritten Digit Recognition and performance of various modelsation[autosaved]SubhradeepMaji
 
Image Segmentation Using Deep Learning : A survey
Image Segmentation Using Deep Learning : A surveyImage Segmentation Using Deep Learning : A survey
Image Segmentation Using Deep Learning : A surveyNUPUR YADAV
 
Transfer Learning (20230516)
Transfer Learning (20230516)Transfer Learning (20230516)
Transfer Learning (20230516)FEG
 
Block modeling using surpac
Block modeling using surpacBlock modeling using surpac
Block modeling using surpacSafdar Ali
 
mode training Disney discussion of vernier.pptx
mode training Disney discussion of vernier.pptxmode training Disney discussion of vernier.pptx
mode training Disney discussion of vernier.pptxhamzaalkhairi802
 
Geofencing_neuralnetworks.pptx
Geofencing_neuralnetworks.pptxGeofencing_neuralnetworks.pptx
Geofencing_neuralnetworks.pptxGeorge John
 
Deep learning fundamental and Research project on IBM POWER9 system from NUS
Deep learning fundamental and Research project on IBM POWER9 system from NUSDeep learning fundamental and Research project on IBM POWER9 system from NUS
Deep learning fundamental and Research project on IBM POWER9 system from NUSGanesan Narayanasamy
 
Teach a neural network to read handwriting
Teach a neural network to read handwritingTeach a neural network to read handwriting
Teach a neural network to read handwritingVipul Kaushal
 
Guiding through a typical Machine Learning Pipeline
Guiding through a typical Machine Learning PipelineGuiding through a typical Machine Learning Pipeline
Guiding through a typical Machine Learning PipelineMichael Gerke
 
Automatic Fabric Defect Detection with a Wide-And-Compact Network
Automatic Fabric Defect Detection with a Wide-And-Compact NetworkAutomatic Fabric Defect Detection with a Wide-And-Compact Network
Automatic Fabric Defect Detection with a Wide-And-Compact NetworkTarik Reza Toha
 
Automated_attendance_system_project.pptx
Automated_attendance_system_project.pptxAutomated_attendance_system_project.pptx
Automated_attendance_system_project.pptxNaveensai51
 
SAMPLING DESIGN AND STEPS IN SAMPLE DESIGN
SAMPLING DESIGN AND STEPS IN SAMPLE DESIGNSAMPLING DESIGN AND STEPS IN SAMPLE DESIGN
SAMPLING DESIGN AND STEPS IN SAMPLE DESIGNpra098
 
Network Intrusion Detection System Using Machine Learning and Deep Learning F...
Network Intrusion Detection System Using Machine Learning and Deep Learning F...Network Intrusion Detection System Using Machine Learning and Deep Learning F...
Network Intrusion Detection System Using Machine Learning and Deep Learning F...Leaving A Legacy
 
Deep Learning and Texture Mapping
Deep Learning and Texture MappingDeep Learning and Texture Mapping
Deep Learning and Texture MappingEfe Kaptan
 
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 digitsNVIDIA Taiwan
 
Malaria Detection Deep Learning Models.pptx
Malaria Detection Deep Learning Models.pptxMalaria Detection Deep Learning Models.pptx
Malaria Detection Deep Learning Models.pptxEmanuel Baisire
 
“A Survey of Model Compression Methods,” a Presentation from Instrumental
“A Survey of Model Compression Methods,” a Presentation from Instrumental“A Survey of Model Compression Methods,” a Presentation from Instrumental
“A Survey of Model Compression Methods,” a Presentation from InstrumentalEdge AI and Vision Alliance
 
face mask detection ppt66 (2).pptx
face mask detection ppt66 (2).pptxface mask detection ppt66 (2).pptx
face mask detection ppt66 (2).pptxShreyaMishra883730
 
Cloud Analytics New Version 2017
Cloud Analytics New Version 2017Cloud Analytics New Version 2017
Cloud Analytics New Version 2017Ajith Kumar Ravi
 

Similar to SPOT THE MASK.pptx (20)

Moving object detection in complex scene
Moving object detection in complex sceneMoving object detection in complex scene
Moving object detection in complex scene
 
Handwritten Digit Recognition and performance of various modelsation[autosaved]
Handwritten Digit Recognition and performance of various modelsation[autosaved]Handwritten Digit Recognition and performance of various modelsation[autosaved]
Handwritten Digit Recognition and performance of various modelsation[autosaved]
 
Image Segmentation Using Deep Learning : A survey
Image Segmentation Using Deep Learning : A surveyImage Segmentation Using Deep Learning : A survey
Image Segmentation Using Deep Learning : A survey
 
Transfer Learning (20230516)
Transfer Learning (20230516)Transfer Learning (20230516)
Transfer Learning (20230516)
 
Block modeling using surpac
Block modeling using surpacBlock modeling using surpac
Block modeling using surpac
 
mode training Disney discussion of vernier.pptx
mode training Disney discussion of vernier.pptxmode training Disney discussion of vernier.pptx
mode training Disney discussion of vernier.pptx
 
Geofencing_neuralnetworks.pptx
Geofencing_neuralnetworks.pptxGeofencing_neuralnetworks.pptx
Geofencing_neuralnetworks.pptx
 
Deep learning fundamental and Research project on IBM POWER9 system from NUS
Deep learning fundamental and Research project on IBM POWER9 system from NUSDeep learning fundamental and Research project on IBM POWER9 system from NUS
Deep learning fundamental and Research project on IBM POWER9 system from NUS
 
Teach a neural network to read handwriting
Teach a neural network to read handwritingTeach a neural network to read handwriting
Teach a neural network to read handwriting
 
Guiding through a typical Machine Learning Pipeline
Guiding through a typical Machine Learning PipelineGuiding through a typical Machine Learning Pipeline
Guiding through a typical Machine Learning Pipeline
 
Automatic Fabric Defect Detection with a Wide-And-Compact Network
Automatic Fabric Defect Detection with a Wide-And-Compact NetworkAutomatic Fabric Defect Detection with a Wide-And-Compact Network
Automatic Fabric Defect Detection with a Wide-And-Compact Network
 
Automated_attendance_system_project.pptx
Automated_attendance_system_project.pptxAutomated_attendance_system_project.pptx
Automated_attendance_system_project.pptx
 
SAMPLING DESIGN AND STEPS IN SAMPLE DESIGN
SAMPLING DESIGN AND STEPS IN SAMPLE DESIGNSAMPLING DESIGN AND STEPS IN SAMPLE DESIGN
SAMPLING DESIGN AND STEPS IN SAMPLE DESIGN
 
Network Intrusion Detection System Using Machine Learning and Deep Learning F...
Network Intrusion Detection System Using Machine Learning and Deep Learning F...Network Intrusion Detection System Using Machine Learning and Deep Learning F...
Network Intrusion Detection System Using Machine Learning and Deep Learning F...
 
Deep Learning and Texture Mapping
Deep Learning and Texture MappingDeep Learning and Texture Mapping
Deep Learning and Texture Mapping
 
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
 
Malaria Detection Deep Learning Models.pptx
Malaria Detection Deep Learning Models.pptxMalaria Detection Deep Learning Models.pptx
Malaria Detection Deep Learning Models.pptx
 
“A Survey of Model Compression Methods,” a Presentation from Instrumental
“A Survey of Model Compression Methods,” a Presentation from Instrumental“A Survey of Model Compression Methods,” a Presentation from Instrumental
“A Survey of Model Compression Methods,” a Presentation from Instrumental
 
face mask detection ppt66 (2).pptx
face mask detection ppt66 (2).pptxface mask detection ppt66 (2).pptx
face mask detection ppt66 (2).pptx
 
Cloud Analytics New Version 2017
Cloud Analytics New Version 2017Cloud Analytics New Version 2017
Cloud Analytics New Version 2017
 

Recently uploaded

haash.tech web design and development company.pdf
haash.tech web design and development company.pdfhaash.tech web design and development company.pdf
haash.tech web design and development company.pdfthahseemE
 
open educational resources power point 3
open educational resources power point 3open educational resources power point 3
open educational resources power point 3brioneil
 
DRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptx
DRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptxDRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptx
DRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptxpostaSimo
 
real estate Image editing hdr, flambient
real estate Image editing hdr, flambientreal estate Image editing hdr, flambient
real estate Image editing hdr, flambientPhotography
 
Diploma Certificate Attestation Service in UAE
Diploma Certificate Attestation Service in UAEDiploma Certificate Attestation Service in UAE
Diploma Certificate Attestation Service in UAEAttestation On Time
 
Office Chairs | Highmoon Office Furniture | Best Quality Ergonomic Chairs
Office Chairs | Highmoon Office Furniture | Best Quality Ergonomic ChairsOffice Chairs | Highmoon Office Furniture | Best Quality Ergonomic Chairs
Office Chairs | Highmoon Office Furniture | Best Quality Ergonomic Chairsakhiladhaneesh5272
 
Best Bitcoin Alternative by Simplyfynews
Best Bitcoin Alternative by SimplyfynewsBest Bitcoin Alternative by Simplyfynews
Best Bitcoin Alternative by SimplyfynewsSimplyfy
 
NEWMAT Stretch Ceilings Service Provider in India
NEWMAT Stretch Ceilings Service Provider in IndiaNEWMAT Stretch Ceilings Service Provider in India
NEWMAT Stretch Ceilings Service Provider in IndiaNEWMATIndia
 
Black box penetration testing: use and process
Black box penetration testing: use and processBlack box penetration testing: use and process
Black box penetration testing: use and processQualysec Technologies
 
Key Pharmaceutical Industry Trends to Watch by 2025
Key Pharmaceutical Industry Trends to Watch by 2025Key Pharmaceutical Industry Trends to Watch by 2025
Key Pharmaceutical Industry Trends to Watch by 2025RNayak3
 
Plumbing Repair: Ensuring Your Home's Comfort and Safety
Plumbing Repair: Ensuring Your Home's Comfort and SafetyPlumbing Repair: Ensuring Your Home's Comfort and Safety
Plumbing Repair: Ensuring Your Home's Comfort and SafetyEco 1 Plumbing LLC
 
Emergency Plumber Services for Repairing and Fitting in Dubai.pdf
Emergency Plumber Services for Repairing and Fitting in Dubai.pdfEmergency Plumber Services for Repairing and Fitting in Dubai.pdf
Emergency Plumber Services for Repairing and Fitting in Dubai.pdfHandyMan Services Dubai
 
Is it profitable to open a financial services business.pptx
Is it profitable to open a financial services business.pptxIs it profitable to open a financial services business.pptx
Is it profitable to open a financial services business.pptxLakshay Gandhi
 
Waste management idea pitch deck v1.pptx
Waste management idea pitch deck v1.pptxWaste management idea pitch deck v1.pptx
Waste management idea pitch deck v1.pptxMarwan Semsom
 
FBI Background Check ! World Class Notary
FBI Background Check ! World Class NotaryFBI Background Check ! World Class Notary
FBI Background Check ! World Class NotaryWorld Class Notary
 

Recently uploaded (19)

NevaClad Refresh_Tellerline Slide Deck2.pdf
NevaClad Refresh_Tellerline Slide Deck2.pdfNevaClad Refresh_Tellerline Slide Deck2.pdf
NevaClad Refresh_Tellerline Slide Deck2.pdf
 
NevaClad Refresh_Tellerline Slide Deck.pdf
NevaClad Refresh_Tellerline Slide Deck.pdfNevaClad Refresh_Tellerline Slide Deck.pdf
NevaClad Refresh_Tellerline Slide Deck.pdf
 
haash.tech web design and development company.pdf
haash.tech web design and development company.pdfhaash.tech web design and development company.pdf
haash.tech web design and development company.pdf
 
open educational resources power point 3
open educational resources power point 3open educational resources power point 3
open educational resources power point 3
 
DRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptx
DRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptxDRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptx
DRAICCHIO slide_Ergonomo_16maggio2024 slides 2.pptx
 
real estate Image editing hdr, flambient
real estate Image editing hdr, flambientreal estate Image editing hdr, flambient
real estate Image editing hdr, flambient
 
4 Reasons Behind the Toilet Blockage.pdf
4 Reasons Behind the Toilet Blockage.pdf4 Reasons Behind the Toilet Blockage.pdf
4 Reasons Behind the Toilet Blockage.pdf
 
Diploma Certificate Attestation Service in UAE
Diploma Certificate Attestation Service in UAEDiploma Certificate Attestation Service in UAE
Diploma Certificate Attestation Service in UAE
 
Office Chairs | Highmoon Office Furniture | Best Quality Ergonomic Chairs
Office Chairs | Highmoon Office Furniture | Best Quality Ergonomic ChairsOffice Chairs | Highmoon Office Furniture | Best Quality Ergonomic Chairs
Office Chairs | Highmoon Office Furniture | Best Quality Ergonomic Chairs
 
Best Bitcoin Alternative by Simplyfynews
Best Bitcoin Alternative by SimplyfynewsBest Bitcoin Alternative by Simplyfynews
Best Bitcoin Alternative by Simplyfynews
 
NEWMAT Stretch Ceilings Service Provider in India
NEWMAT Stretch Ceilings Service Provider in IndiaNEWMAT Stretch Ceilings Service Provider in India
NEWMAT Stretch Ceilings Service Provider in India
 
Black box penetration testing: use and process
Black box penetration testing: use and processBlack box penetration testing: use and process
Black box penetration testing: use and process
 
Key Pharmaceutical Industry Trends to Watch by 2025
Key Pharmaceutical Industry Trends to Watch by 2025Key Pharmaceutical Industry Trends to Watch by 2025
Key Pharmaceutical Industry Trends to Watch by 2025
 
Plumbing Repair: Ensuring Your Home's Comfort and Safety
Plumbing Repair: Ensuring Your Home's Comfort and SafetyPlumbing Repair: Ensuring Your Home's Comfort and Safety
Plumbing Repair: Ensuring Your Home's Comfort and Safety
 
Emergency Plumber Services for Repairing and Fitting in Dubai.pdf
Emergency Plumber Services for Repairing and Fitting in Dubai.pdfEmergency Plumber Services for Repairing and Fitting in Dubai.pdf
Emergency Plumber Services for Repairing and Fitting in Dubai.pdf
 
Is it profitable to open a financial services business.pptx
Is it profitable to open a financial services business.pptxIs it profitable to open a financial services business.pptx
Is it profitable to open a financial services business.pptx
 
Rainwater Harvesting System .
Rainwater Harvesting System                        .Rainwater Harvesting System                        .
Rainwater Harvesting System .
 
Waste management idea pitch deck v1.pptx
Waste management idea pitch deck v1.pptxWaste management idea pitch deck v1.pptx
Waste management idea pitch deck v1.pptx
 
FBI Background Check ! World Class Notary
FBI Background Check ! World Class NotaryFBI Background Check ! World Class Notary
FBI Background Check ! World Class Notary
 

SPOT THE MASK.pptx

  • 2. INTRODUCTION • This challenge involves employing Computer Vision to identify whether a person is wearing nose masks or not. • This challenge is to help ensure that masks are being worn appropriately in public. • The goal is to use data to create an image classification model to accurately predict whether an image contains a person wearing nose mask or not 20XX presentation title 2
  • 3. MODEL PIPELINE • DATA STRUCTURE ANALYSIS • DATA PREPROCESSING • DATA PREPARATION • MODEL SELECTION • MODEL TRAINING • MODEL AND HYPERPARAMTER FINE-TUNING • MODEL EVALUATION 20XX presentation title 3
  • 4. METHODOLOGY(1) • DATA STRUCTURE ANALYSIS • USE OF PANDA DATAFRAME FUNCTIONS TO ANALYZE THE DATA FROM THE TRAIN.CSV • TRAIN AND TEST IMAGES BOTH GROUPED INTO ONE FOLDER • DATA PREPROCESSING • TRAIN AND TEST IMAGES BOTH GROUPED INTO ONE FOLDER HAD TO BE SEPARATE FOLDERS • USING THE TARGET IN THE TRAIN..CSV FILE, THE IMAGES WERE SEPARATED INTO WITH NOSE MASKS AND WITHOUT NOSE MASKS • DATA PREPARATION • PREPROCESS THE DATA BY RESIZING THE IMAGES AND CREATED DATAFRAMES FOR BOTH WITH NOSE MASKS AND WITHOUT NOSE MASKS • USE AN IMAGE DATA GENERATOR FROM TENSORFLOW TO AUGMENT THE DATA • SPLIT THE DATASET AFTER AUGMENTATION INTO TWO USING TRAIN_TEST_SPLIT 20XX presentation title 4
  • 5. METHODOLOGY(2) • MODEL SELECTION • PRE-TRAINED MODELS SUCH AS VGG16, RESNET AND DENSET WERE SELECTED FOR THEIR IMAGE CLASSIFICATION ABILITIES • A CONVOLUTIONAL NEURAL NETWORK (CNN) IS BUILT USING THE PRE TRAINED MODELS • MODEL TRAINING • USING THE PRE-TRAINED MODEL (DENSENET200), A CNN OF 3 LAYERS IS BUILT WITH GLOBALAVERAGEPOOLING2D, DENSE AND DROPOUT • THE MODEL IS TRAINED USING THE DATASET WITH ONE SET TRAIINING AND THE OTHER SET EVALUATING IT • MODEL AND HYPERPARAMETERS FINE-TUNING • THE MODEL IS FINE-TUNED BY FREEZING SOME LAYERS AND TRAINING THE REMAINING LAYERS • HYPERPARAMETERS ARE OPTIMIZED AS WELL TO IMPROVE PERFORMANCE 20XX presentation title 5
  • 6. METHODOLOGY(3) • MODEL EVALUATION • TEST MODEL ON THE TEST DATASET PROVIDED TO ENSURE MODEL IS WORKING 20XX presentation title 6
  • 7. RESULTS • THE CNN MODEL YIELED AN ACCURACY OF 0.95875 DURING TRAINING AND IMPROVED TO 0.98769 • THE ACCURACY OF THE TEST WAS 0.990180639 20XX presentation title 7