SlideShare a Scribd company logo
Role of AI & ML in
Beauty Care Industry
GeekNight :: 14th November, 2018
Piyush Bhargava | (Hyderabad)
Agenda
● What is AI & ML ?
● AI & ML in Beauty Care
● Examples
● Skin Analysis
● Hair Analysis
● Virtual Makeup
● Live demo of Virtual Makeup
● Q&A
2
Artificial Intelligence vs Machine Learning ?
● AI is not a system, but a system
has AI
● ML is one of the ways to make a
system AI enabled
● ML is making a machine learn
○ from Experience (E) - Training Data
○ w.r.t some Task (T) - Problem
Statement
○ and a Performance Measure (P) -
Accuracy
when P improves with E
Artificial Intelligence
Machine Learning
Deep Learning
3
AI & ML in Beauty Care
● Product Recommendation
● Product Personalisation
● Product Validation & Research
● Retail Experiences
○ Virtual (Magic) Mirrors
○ Virtual Try On Apps (Web & Mobile)
● Constant Learning AI Systems
4
Some big p:)ayers | Unique So:)utions
Recommendation Engines
● Proven - Personalised Skin Care
● My Beauty Matches - Personalised Product Recommendation
● Function of Beauty - Personalised Hair Care
● Beauty.ai - Deep Learning based Beauty Contest
● Olay - Skin analysis from Selfie
Augmented Reality
● Modiface - SDKs for Beauty and Face, AR Mirror
● Perfect Corp. - Virtual Makeup Mobile app
● Coty - Magic Mirror for Bourjois Velvet Collection
5
Some big p:)ayers | Unique So:)utions
6Video Source: https://www.coty.com/in-the-news/press-release/coty-magic-mirror
Product Validation
● Objective - Improve product performance
● Experimental Cosmetics tested on subjects
● Skin / Hair analysis performed through ML / CV modules
● Based on the analysis, ingredients are finalised for the specific
product
● At product launch, statistical data used to back the performance
claims
7
Product Recommendation
8
Quiz / Customer
Profiling
Matching
Product
Recommendations
A sample product
markup table
Visitors profile
estimation
Skin Analysis
● Skin Advisor Applications
● Mobile Skin & Hair Analysis Devices
○ Octa Core Processor
○ 30 to 500 times high magnification lens
○ Normal and UV light sensors
○ Android based, WiFi mirroring
● Pore Analysis
● Moisture / Oiliness
● Wrinkles
● Melanin (Color) / Pigmentation
● Sebum / Acne (T & U Zone)
● Morphological Operations, Edge and Contour
analysis, Color based Segmentation, GLCM
9Image Source: http://www.aramhuvis.com/en/apm/
Skin Analysis :: Machine Learning Approach
10
Used for detection of symptoms and sometimes
even categorisation of symptoms based on
severity (high, medium, low)
● Crow’s feet
● Blackheads
● Acne
● Roughness / Dryness
● Eye bags
Image
Pre-processing
RoI Selection
Features Extraction
Classifier Training
Feature Vector
Classifier Model
Detection /
Recognition
Skin Color Analysis using ICA
Such a tool helps in
● Studying the current
pigmentation composition of a
subject’s skin
● Recommending specific skin
creams based on the analysis
above
● Simulate the proposed
improvements in skin through
reconstruction
● Observe and monitor the actual
changes over a period of time
11
Image
ICA
Haemoglobin (H) Melanin (M)
Modified (H’) Modified (M’)
Inverse
ICA/PCA
PCA
Reconstructed
Image
Hair / Scalp Surface Analysis
● Hair Cuticle Damage Analysis
○ Uplifted Cuticle
○ Chipped Cuticle
○ Missing Cuticle
● Broken and Split ends detection
● Hair Density Analysis
● Sebum / Acne analysis on Scalp Surface
● High Magnification Mobile Device used for
capturing the Hair / Scalp images for further
processing
● The tool helps in assessing the hair and scalp
surface, identifying a health index for hair and
recommend the most suitable Shampoo and
Conditioner (and even hair oil)
12
Image source: https://www.youtube.com/watch?v=FmSkffcqBxY&t=700s
Image source: https://www.cgtrader.com/3d-models/science/medical/damaged-hair-cuticle-layer-and-healthy-hair-cuticle-layer
Virtual Try On / Makeup
● Simulating makeup on self
● Google Search for “Virtual Makeup” shows many online portals
offering such services
● All major cosmetic brands now offer Virtual Makeup feature on
their website and mobile apps
○ Try makeup on Eyes, Lips and Face, Hair Color
○ Choose from different shades
○ Add product to cart
○ Order online or collect from nearest store
● Convenient and Comfortable
● More choices than a “brick and mortar” shop
● A better visualization of products on self
● Look Good, Feel Better factor increases popularity among common
people
● Steady source of revenue for companies like Modiface, Taaz,
Perfect365, etc offering Virtual Makeover Platform / Engine as a
product to many cosmetic companies
13
Image Source: http://www.flickr.com/photos/idhren/6236954553/
Shape Predictor / Face Landmark Extraction
● Shape predictor is used in analysis of
structures and sub regions of an object
● For faces, landmark points describe the
shape of the face attributes like: eyes,
eyebrows, nose, mouth, jawline & chin
● Wide range of applications, including:
face recognition, face filters (snapchat
style), face morphing, emotion
recognition, blink detection, ...
● Standard pre-trained detectors available
● New shape predictors can be trained
● New dataset can also be generated
68 landmark points generated by Dlib
14
Training a Shape Predictor using Dlib
● Create or Pick/Download a large dataset of images
○ Mirroring and Augmentation increase the dataset size further
● Generate landmark annotations (.xml file)
○ Face region :: <top, left, width, height>
○ N points :: <x, y> where N is the number of landmarks we want to train for
● CNN training options affect Size, Speed and Accuracy
● Manually fine-tuned in order to get the desired performance
● Face detection (like OpenCV Haar Cascade, Dlib HOG Detector, CNN detectors)
● The points must be
○ Manually labelled
○ Manually reviewed and corrected
● Example python code (http://dlib.net/train_shape_predictor.py.html)
● Example c++ code (http://dlib.net/train_shape_predictor_ex.cpp.html)
● Also, there is a blog post on Medium that explains training a shape predictor using Dlib
15
A generic pipeline for Virtual Makeup
1. Face Detection
a. (HAAR, SVM+HOG detector, CNN)
2. Facial Landmark Points extraction
a. (using DLib)
3. Lip Mask Creation through Interpolation
and Curve Fitting
4. Overlaying Lipstick color through Alpha
Compositing
5. Refinements (for more realistic effects)
6. Similar stages for other makeup like eye
liner, eye shadow, etc.
16
Facial Landmark Points Extraction using DLib
● Deep Learning based
implementation
● Can be trained for custom
number of landmark points
● Real time, good for Video / Live
processing
● Landmark points help in
○ Head Pose estimation,
○ Facial Shape analysis
○ Facial Region Extraction
and Analysis
○ Many more
Image source: https://ibug.doc.ic.ac.uk/resources/facial-point-annotations/ 17
Lip Mask Creation
Normal
Smiling
Open Mouth
Lip Contour
Lip landmark points
18
Extract lip points
Interpolate to get smoother
lip contour
Create a Mask
Blur the Mask
68 landmark points
Apply Lipstick
Alpha Compositing - Augmenting on Live Frame
● Alpha Blending ensures
○ Smoother edges
○ Lesser Artifacts
○ Transparency effect to retain
texture from original image
+
HSV Colored Lip
Region
Blurred Mask for
Lip Region
Without Alpha
Blending
With Alpha
Blending
19
Let’s have some fun :)
Choose a Lipstick shade
Choose a shade for Eyeshadow
Virtual Makeup in action ;)
20
Cha:((enges & Future Landscape
● Data
● Constant innovation to stay ahead of the Competition
● Voice based systems
● Rendering Realistic Effects
○ Using Skin Surface Reflective Properties
○ Using Ambient Lighting Information
○ Compute intensive
● Shift towards Natural and Organic Products
● Personalised Designing & Packaging of Products
● Shifting lab based cosmetic tests to home environments
○ Using cheaper but reliable sensors
○ Increases the data sources
21
Acknowledgements
● Adrian Rosebrock for online tutorials at PyImageSearch
● Intelligent Behavior Understanding Group (iBUG) for providing
annotated datasets
● Luca Anzalone for his Medium blog on Shape Predictor Training
● REES46 for their Medium blog post on Progressive Personalization
● ThoughtWorks Hyderabad
● The Awesome Audience .. that’s You !! :)
22
Q&A
23
Thank You !!
Piyush Bhargava | (Hyderabad)
piyush.bhargava@thoughtworks.com

More Related Content

What's hot

artificial intelligence
artificial intelligence artificial intelligence
artificial intelligence Ali Raza
 
Artificial Intelligence power point presentation
Artificial Intelligence power point presentationArtificial Intelligence power point presentation
Artificial Intelligence power point presentationDavid Raj Kanthi
 
Introduction to Artificial Intelligence
Introduction to Artificial IntelligenceIntroduction to Artificial Intelligence
Introduction to Artificial Intelligencesnehal_152
 
Artificial Intelligence (A.I) and Its Application -Seminar
Artificial Intelligence (A.I) and Its Application -SeminarArtificial Intelligence (A.I) and Its Application -Seminar
Artificial Intelligence (A.I) and Its Application -SeminarBIJAY NAYAK
 
artificial intelligence
artificial intelligenceartificial intelligence
artificial intelligencevallibhargavi
 
The future of ai ethics
The future of ai ethics The future of ai ethics
The future of ai ethics Vish Nandlall
 
Artificial Intelligence and Robotics
Artificial Intelligence and RoboticsArtificial Intelligence and Robotics
Artificial Intelligence and Roboticsvijayrock442
 
Face Recognition Technology
Face Recognition TechnologyFace Recognition Technology
Face Recognition TechnologyAgrani Rastogi
 
Artificial Intelligence PPT
Artificial Intelligence PPTArtificial Intelligence PPT
Artificial Intelligence PPTSachin Sakure
 
Artificial Intelligence: Role In Pharma Sector
Artificial Intelligence: Role In Pharma SectorArtificial Intelligence: Role In Pharma Sector
Artificial Intelligence: Role In Pharma SectorJayBhavsar41
 
Artificial Intelligence: Predictions for 2017
Artificial Intelligence: Predictions for 2017Artificial Intelligence: Predictions for 2017
Artificial Intelligence: Predictions for 2017NVIDIA
 
HUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORK
HUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORKHUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORK
HUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORKijiert bestjournal
 
Artificial intelligence - An Overview
Artificial intelligence - An OverviewArtificial intelligence - An Overview
Artificial intelligence - An OverviewGiri Dharan
 
Artificial Intelligence
Artificial Intelligence Artificial Intelligence
Artificial Intelligence Midhun
 

What's hot (20)

artificial intelligence
artificial intelligence artificial intelligence
artificial intelligence
 
Artificial Intelligence power point presentation
Artificial Intelligence power point presentationArtificial Intelligence power point presentation
Artificial Intelligence power point presentation
 
Artificial Intelligence
Artificial IntelligenceArtificial Intelligence
Artificial Intelligence
 
Introduction to Artificial Intelligence
Introduction to Artificial IntelligenceIntroduction to Artificial Intelligence
Introduction to Artificial Intelligence
 
Artificial Intelligence (A.I) and Its Application -Seminar
Artificial Intelligence (A.I) and Its Application -SeminarArtificial Intelligence (A.I) and Its Application -Seminar
Artificial Intelligence (A.I) and Its Application -Seminar
 
artificial intelligence
artificial intelligenceartificial intelligence
artificial intelligence
 
The future of ai ethics
The future of ai ethics The future of ai ethics
The future of ai ethics
 
Artificial Intelligence and Robotics
Artificial Intelligence and RoboticsArtificial Intelligence and Robotics
Artificial Intelligence and Robotics
 
Artificial Intelligence
Artificial IntelligenceArtificial Intelligence
Artificial Intelligence
 
Face Recognition Technology
Face Recognition TechnologyFace Recognition Technology
Face Recognition Technology
 
Artificial intelligence
Artificial intelligenceArtificial intelligence
Artificial intelligence
 
QR code
QR codeQR code
QR code
 
Wrinkled skin
Wrinkled skinWrinkled skin
Wrinkled skin
 
Artificial Intelligence PPT
Artificial Intelligence PPTArtificial Intelligence PPT
Artificial Intelligence PPT
 
Artificial Intelligence: Role In Pharma Sector
Artificial Intelligence: Role In Pharma SectorArtificial Intelligence: Role In Pharma Sector
Artificial Intelligence: Role In Pharma Sector
 
Artificial Intelligence: Predictions for 2017
Artificial Intelligence: Predictions for 2017Artificial Intelligence: Predictions for 2017
Artificial Intelligence: Predictions for 2017
 
HUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORK
HUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORKHUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORK
HUMAN FACE RECOGNITION USING IMAGE PROCESSING PCA AND NEURAL NETWORK
 
Skin care products
Skin care products Skin care products
Skin care products
 
Artificial intelligence - An Overview
Artificial intelligence - An OverviewArtificial intelligence - An Overview
Artificial intelligence - An Overview
 
Artificial Intelligence
Artificial Intelligence Artificial Intelligence
Artificial Intelligence
 

Similar to Role of AI & ML in beauty care industry

2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptx2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptxgdgsurrey
 
May internship challenge: User Authentication System only using image data: C...
May internship challenge: User Authentication System only using image data: C...May internship challenge: User Authentication System only using image data: C...
May internship challenge: User Authentication System only using image data: C...Ridge-i, Inc.
 
Product Culture with Property Finder VP Product
Product Culture with Property Finder VP ProductProduct Culture with Property Finder VP Product
Product Culture with Property Finder VP ProductProduct School
 
Investing in ai driven startups
Investing in ai driven startupsInvesting in ai driven startups
Investing in ai driven startupsRoy Lowrance
 
Project presentation by Debendra Adhikari
Project presentation by Debendra AdhikariProject presentation by Debendra Adhikari
Project presentation by Debendra AdhikariDEBENDRA ADHIKARI
 
Digitalmicroscope(with android)
Digitalmicroscope(with android)Digitalmicroscope(with android)
Digitalmicroscope(with android)Microscope Ryu
 
Facial Recognition Attendance System (Synopsis).pptx
Facial Recognition Attendance System (Synopsis).pptxFacial Recognition Attendance System (Synopsis).pptx
Facial Recognition Attendance System (Synopsis).pptxkakimetu
 
A.I. in the Enterprise: Computer Vision
A.I. in the Enterprise: Computer VisionA.I. in the Enterprise: Computer Vision
A.I. in the Enterprise: Computer VisionChristopher Mohritz
 
Running Successful App Ads Campaigns - Hero Conf 2019 - Greg Young
Running Successful App Ads Campaigns - Hero Conf 2019 - Greg YoungRunning Successful App Ads Campaigns - Hero Conf 2019 - Greg Young
Running Successful App Ads Campaigns - Hero Conf 2019 - Greg YoungGreg Young
 
Face detection using template matching
Face detection using template matchingFace detection using template matching
Face detection using template matchingBrijesh Borad
 
Face and Eye Detection Varying Scenarios With Haar Classifier_2015
Face and Eye Detection Varying Scenarios With Haar Classifier_2015Face and Eye Detection Varying Scenarios With Haar Classifier_2015
Face and Eye Detection Varying Scenarios With Haar Classifier_2015Showrav Mazumder
 
Webinar - Know Your Customer - Arya (20160526)
Webinar - Know Your Customer - Arya (20160526)Webinar - Know Your Customer - Arya (20160526)
Webinar - Know Your Customer - Arya (20160526)Turi, Inc.
 
face-recognition-technology-ppt[1].pptx
face-recognition-technology-ppt[1].pptxface-recognition-technology-ppt[1].pptx
face-recognition-technology-ppt[1].pptxTanayChakraborty11
 
Intermediate Presentation (1)
Intermediate Presentation (1)Intermediate Presentation (1)
Intermediate Presentation (1)Justin Hernandez
 
Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...
Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...
Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...Christopher Mohritz
 
AI and ML for Product Management by Smartsheet Sr Dir of PM
AI and ML for Product Management by Smartsheet Sr Dir of PMAI and ML for Product Management by Smartsheet Sr Dir of PM
AI and ML for Product Management by Smartsheet Sr Dir of PMProduct School
 

Similar to Role of AI & ML in beauty care industry (20)

2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptx2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptx
 
May internship challenge: User Authentication System only using image data: C...
May internship challenge: User Authentication System only using image data: C...May internship challenge: User Authentication System only using image data: C...
May internship challenge: User Authentication System only using image data: C...
 
Karakuri
KarakuriKarakuri
Karakuri
 
Pattern recognition
Pattern recognitionPattern recognition
Pattern recognition
 
Product Culture with Property Finder VP Product
Product Culture with Property Finder VP ProductProduct Culture with Property Finder VP Product
Product Culture with Property Finder VP Product
 
AI & Programmatic Advertising Master Class - Yang Han, StackAdapt
AI & Programmatic Advertising Master Class - Yang Han, StackAdaptAI & Programmatic Advertising Master Class - Yang Han, StackAdapt
AI & Programmatic Advertising Master Class - Yang Han, StackAdapt
 
Investing in ai driven startups
Investing in ai driven startupsInvesting in ai driven startups
Investing in ai driven startups
 
Project presentation by Debendra Adhikari
Project presentation by Debendra AdhikariProject presentation by Debendra Adhikari
Project presentation by Debendra Adhikari
 
Digitalmicroscope(with android)
Digitalmicroscope(with android)Digitalmicroscope(with android)
Digitalmicroscope(with android)
 
Facial Recognition Attendance System (Synopsis).pptx
Facial Recognition Attendance System (Synopsis).pptxFacial Recognition Attendance System (Synopsis).pptx
Facial Recognition Attendance System (Synopsis).pptx
 
A.I. in the Enterprise: Computer Vision
A.I. in the Enterprise: Computer VisionA.I. in the Enterprise: Computer Vision
A.I. in the Enterprise: Computer Vision
 
Running Successful App Ads Campaigns - Hero Conf 2019 - Greg Young
Running Successful App Ads Campaigns - Hero Conf 2019 - Greg YoungRunning Successful App Ads Campaigns - Hero Conf 2019 - Greg Young
Running Successful App Ads Campaigns - Hero Conf 2019 - Greg Young
 
Face detection using template matching
Face detection using template matchingFace detection using template matching
Face detection using template matching
 
Face and Eye Detection Varying Scenarios With Haar Classifier_2015
Face and Eye Detection Varying Scenarios With Haar Classifier_2015Face and Eye Detection Varying Scenarios With Haar Classifier_2015
Face and Eye Detection Varying Scenarios With Haar Classifier_2015
 
Webinar - Know Your Customer - Arya (20160526)
Webinar - Know Your Customer - Arya (20160526)Webinar - Know Your Customer - Arya (20160526)
Webinar - Know Your Customer - Arya (20160526)
 
face-recognition-technology-ppt[1].pptx
face-recognition-technology-ppt[1].pptxface-recognition-technology-ppt[1].pptx
face-recognition-technology-ppt[1].pptx
 
Salesforce Einstein: Use Cases and Product Features
Salesforce Einstein: Use Cases and Product FeaturesSalesforce Einstein: Use Cases and Product Features
Salesforce Einstein: Use Cases and Product Features
 
Intermediate Presentation (1)
Intermediate Presentation (1)Intermediate Presentation (1)
Intermediate Presentation (1)
 
Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...
Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...
Computers Are Opening Their Eyes - And They're Already Better at Seeing Than ...
 
AI and ML for Product Management by Smartsheet Sr Dir of PM
AI and ML for Product Management by Smartsheet Sr Dir of PMAI and ML for Product Management by Smartsheet Sr Dir of PM
AI and ML for Product Management by Smartsheet Sr Dir of PM
 

More from GeekNightHyderabad

Testing strategies in microservices
Testing strategies in microservicesTesting strategies in microservices
Testing strategies in microservicesGeekNightHyderabad
 
Scaling enterprise digital platforms with kubernetes
Scaling enterprise digital platforms with kubernetesScaling enterprise digital platforms with kubernetes
Scaling enterprise digital platforms with kubernetesGeekNightHyderabad
 
FreedomBox & Community Wi-Fi networks
FreedomBox & Community Wi-Fi networksFreedomBox & Community Wi-Fi networks
FreedomBox & Community Wi-Fi networksGeekNightHyderabad
 
Rendezvous with aucovei (autonomous connected car)
Rendezvous with aucovei (autonomous connected car)Rendezvous with aucovei (autonomous connected car)
Rendezvous with aucovei (autonomous connected car)GeekNightHyderabad
 
Design lean agile_thinking presentation
Design lean agile_thinking presentationDesign lean agile_thinking presentation
Design lean agile_thinking presentationGeekNightHyderabad
 
Hardware hacking and internet of things
Hardware hacking and internet of thingsHardware hacking and internet of things
Hardware hacking and internet of thingsGeekNightHyderabad
 
Spring to Cloud - REST To Microservices
Spring to Cloud - REST To MicroservicesSpring to Cloud - REST To Microservices
Spring to Cloud - REST To MicroservicesGeekNightHyderabad
 
Building Cloud Native Applications Using Spring Boot and Spring Cloud
Building Cloud Native Applications Using Spring Boot and Spring CloudBuilding Cloud Native Applications Using Spring Boot and Spring Cloud
Building Cloud Native Applications Using Spring Boot and Spring CloudGeekNightHyderabad
 
Progressive Web Applications - The Next Gen Web Technologies
Progressive Web Applications - The Next Gen Web TechnologiesProgressive Web Applications - The Next Gen Web Technologies
Progressive Web Applications - The Next Gen Web TechnologiesGeekNightHyderabad
 
Scaling a Game Server: From 500 to 100,000 Users
Scaling a Game Server: From 500 to 100,000 UsersScaling a Game Server: From 500 to 100,000 Users
Scaling a Game Server: From 500 to 100,000 UsersGeekNightHyderabad
 
Big Data - Need of Converged Data Platform
Big Data - Need of Converged Data PlatformBig Data - Need of Converged Data Platform
Big Data - Need of Converged Data PlatformGeekNightHyderabad
 
Building a Data Lake - An App Dev's Perspective
Building a Data Lake - An App Dev's PerspectiveBuilding a Data Lake - An App Dev's Perspective
Building a Data Lake - An App Dev's PerspectiveGeekNightHyderabad
 
Understanding the Intelligent Cloud
Understanding the Intelligent CloudUnderstanding the Intelligent Cloud
Understanding the Intelligent CloudGeekNightHyderabad
 
GeekNight 22.0 Multi-paradigm programming in Scala and Akka
GeekNight 22.0 Multi-paradigm programming in Scala and AkkaGeekNight 22.0 Multi-paradigm programming in Scala and Akka
GeekNight 22.0 Multi-paradigm programming in Scala and AkkaGeekNightHyderabad
 

More from GeekNightHyderabad (20)

Testing strategies in microservices
Testing strategies in microservicesTesting strategies in microservices
Testing strategies in microservices
 
Metaprogramming ruby
Metaprogramming rubyMetaprogramming ruby
Metaprogramming ruby
 
Scaling enterprise digital platforms with kubernetes
Scaling enterprise digital platforms with kubernetesScaling enterprise digital platforms with kubernetes
Scaling enterprise digital platforms with kubernetes
 
FreedomBox & Community Wi-Fi networks
FreedomBox & Community Wi-Fi networksFreedomBox & Community Wi-Fi networks
FreedomBox & Community Wi-Fi networks
 
Rendezvous with aucovei (autonomous connected car)
Rendezvous with aucovei (autonomous connected car)Rendezvous with aucovei (autonomous connected car)
Rendezvous with aucovei (autonomous connected car)
 
Breaking down a monolith
Breaking down a monolithBreaking down a monolith
Breaking down a monolith
 
Design lean agile_thinking presentation
Design lean agile_thinking presentationDesign lean agile_thinking presentation
Design lean agile_thinking presentation
 
Scaling pipelines
Scaling pipelinesScaling pipelines
Scaling pipelines
 
Blockchain beyond bitcoin
Blockchain beyond bitcoinBlockchain beyond bitcoin
Blockchain beyond bitcoin
 
Http/2
Http/2Http/2
Http/2
 
Hardware hacking and internet of things
Hardware hacking and internet of thingsHardware hacking and internet of things
Hardware hacking and internet of things
 
Spring to Cloud - REST To Microservices
Spring to Cloud - REST To MicroservicesSpring to Cloud - REST To Microservices
Spring to Cloud - REST To Microservices
 
Serverless
ServerlessServerless
Serverless
 
Building Cloud Native Applications Using Spring Boot and Spring Cloud
Building Cloud Native Applications Using Spring Boot and Spring CloudBuilding Cloud Native Applications Using Spring Boot and Spring Cloud
Building Cloud Native Applications Using Spring Boot and Spring Cloud
 
Progressive Web Applications - The Next Gen Web Technologies
Progressive Web Applications - The Next Gen Web TechnologiesProgressive Web Applications - The Next Gen Web Technologies
Progressive Web Applications - The Next Gen Web Technologies
 
Scaling a Game Server: From 500 to 100,000 Users
Scaling a Game Server: From 500 to 100,000 UsersScaling a Game Server: From 500 to 100,000 Users
Scaling a Game Server: From 500 to 100,000 Users
 
Big Data - Need of Converged Data Platform
Big Data - Need of Converged Data PlatformBig Data - Need of Converged Data Platform
Big Data - Need of Converged Data Platform
 
Building a Data Lake - An App Dev's Perspective
Building a Data Lake - An App Dev's PerspectiveBuilding a Data Lake - An App Dev's Perspective
Building a Data Lake - An App Dev's Perspective
 
Understanding the Intelligent Cloud
Understanding the Intelligent CloudUnderstanding the Intelligent Cloud
Understanding the Intelligent Cloud
 
GeekNight 22.0 Multi-paradigm programming in Scala and Akka
GeekNight 22.0 Multi-paradigm programming in Scala and AkkaGeekNight 22.0 Multi-paradigm programming in Scala and Akka
GeekNight 22.0 Multi-paradigm programming in Scala and Akka
 

Recently uploaded

EY - Supply Chain Services 2018_template.pptx
EY - Supply Chain Services 2018_template.pptxEY - Supply Chain Services 2018_template.pptx
EY - Supply Chain Services 2018_template.pptxAlguinaldoKong
 
Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...
Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...
Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...Sérgio Sacani
 
word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...
word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...
word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...Subhajit Sahu
 
Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...
Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...
Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...Sérgio Sacani
 
insect taxonomy importance systematics and classification
insect taxonomy importance systematics and classificationinsect taxonomy importance systematics and classification
insect taxonomy importance systematics and classificationanitaento25
 
Plant Biotechnology undergraduates note.pptx
Plant Biotechnology undergraduates note.pptxPlant Biotechnology undergraduates note.pptx
Plant Biotechnology undergraduates note.pptxyusufzako14
 
Large scale production of streptomycin.pptx
Large scale production of streptomycin.pptxLarge scale production of streptomycin.pptx
Large scale production of streptomycin.pptxCherry
 
Structural Classification Of Protein (SCOP)
Structural Classification Of Protein  (SCOP)Structural Classification Of Protein  (SCOP)
Structural Classification Of Protein (SCOP)aishnasrivastava
 
BLOOD AND BLOOD COMPONENT- introduction to blood physiology
BLOOD AND BLOOD COMPONENT- introduction to blood physiologyBLOOD AND BLOOD COMPONENT- introduction to blood physiology
BLOOD AND BLOOD COMPONENT- introduction to blood physiologyNoelManyise1
 
GBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of LipidsGBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of LipidsAreesha Ahmad
 
Cancer cell metabolism: special Reference to Lactate Pathway
Cancer cell metabolism: special Reference to Lactate PathwayCancer cell metabolism: special Reference to Lactate Pathway
Cancer cell metabolism: special Reference to Lactate PathwayAADYARAJPANDEY1
 
Mammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also FunctionsMammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also FunctionsYOGESH DOGRA
 
GEOLOGICAL FIELD REPORT On Kaptai Rangamati Road-Cut Section.pdf
GEOLOGICAL FIELD REPORT  On  Kaptai Rangamati Road-Cut Section.pdfGEOLOGICAL FIELD REPORT  On  Kaptai Rangamati Road-Cut Section.pdf
GEOLOGICAL FIELD REPORT On Kaptai Rangamati Road-Cut Section.pdfUniversity of Barishal
 
PRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATION
PRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATIONPRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATION
PRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATIONChetanK57
 
extra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdfextra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdfDiyaBiswas10
 
NuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final versionNuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final versionpablovgd
 
FAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable PredictionsFAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable PredictionsMichel Dumontier
 
Lab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerinLab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerinossaicprecious19
 
GBSN- Microbiology (Lab 3) Gram Staining
GBSN- Microbiology (Lab 3) Gram StainingGBSN- Microbiology (Lab 3) Gram Staining
GBSN- Microbiology (Lab 3) Gram StainingAreesha Ahmad
 
Seminar of U.V. Spectroscopy by SAMIR PANDA
 Seminar of U.V. Spectroscopy by SAMIR PANDA Seminar of U.V. Spectroscopy by SAMIR PANDA
Seminar of U.V. Spectroscopy by SAMIR PANDASAMIR PANDA
 

Recently uploaded (20)

EY - Supply Chain Services 2018_template.pptx
EY - Supply Chain Services 2018_template.pptxEY - Supply Chain Services 2018_template.pptx
EY - Supply Chain Services 2018_template.pptx
 
Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...
Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...
Earliest Galaxies in the JADES Origins Field: Luminosity Function and Cosmic ...
 
word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...
word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...
word2vec, node2vec, graph2vec, X2vec: Towards a Theory of Vector Embeddings o...
 
Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...
Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...
Gliese 12 b, a temperate Earth-sized planet at 12 parsecs discovered with TES...
 
insect taxonomy importance systematics and classification
insect taxonomy importance systematics and classificationinsect taxonomy importance systematics and classification
insect taxonomy importance systematics and classification
 
Plant Biotechnology undergraduates note.pptx
Plant Biotechnology undergraduates note.pptxPlant Biotechnology undergraduates note.pptx
Plant Biotechnology undergraduates note.pptx
 
Large scale production of streptomycin.pptx
Large scale production of streptomycin.pptxLarge scale production of streptomycin.pptx
Large scale production of streptomycin.pptx
 
Structural Classification Of Protein (SCOP)
Structural Classification Of Protein  (SCOP)Structural Classification Of Protein  (SCOP)
Structural Classification Of Protein (SCOP)
 
BLOOD AND BLOOD COMPONENT- introduction to blood physiology
BLOOD AND BLOOD COMPONENT- introduction to blood physiologyBLOOD AND BLOOD COMPONENT- introduction to blood physiology
BLOOD AND BLOOD COMPONENT- introduction to blood physiology
 
GBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of LipidsGBSN - Biochemistry (Unit 5) Chemistry of Lipids
GBSN - Biochemistry (Unit 5) Chemistry of Lipids
 
Cancer cell metabolism: special Reference to Lactate Pathway
Cancer cell metabolism: special Reference to Lactate PathwayCancer cell metabolism: special Reference to Lactate Pathway
Cancer cell metabolism: special Reference to Lactate Pathway
 
Mammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also FunctionsMammalian Pineal Body Structure and Also Functions
Mammalian Pineal Body Structure and Also Functions
 
GEOLOGICAL FIELD REPORT On Kaptai Rangamati Road-Cut Section.pdf
GEOLOGICAL FIELD REPORT  On  Kaptai Rangamati Road-Cut Section.pdfGEOLOGICAL FIELD REPORT  On  Kaptai Rangamati Road-Cut Section.pdf
GEOLOGICAL FIELD REPORT On Kaptai Rangamati Road-Cut Section.pdf
 
PRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATION
PRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATIONPRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATION
PRESENTATION ABOUT PRINCIPLE OF COSMATIC EVALUATION
 
extra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdfextra-chromosomal-inheritance[1].pptx.pdfpdf
extra-chromosomal-inheritance[1].pptx.pdfpdf
 
NuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final versionNuGOweek 2024 Ghent - programme - final version
NuGOweek 2024 Ghent - programme - final version
 
FAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable PredictionsFAIR & AI Ready KGs for Explainable Predictions
FAIR & AI Ready KGs for Explainable Predictions
 
Lab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerinLab report on liquid viscosity of glycerin
Lab report on liquid viscosity of glycerin
 
GBSN- Microbiology (Lab 3) Gram Staining
GBSN- Microbiology (Lab 3) Gram StainingGBSN- Microbiology (Lab 3) Gram Staining
GBSN- Microbiology (Lab 3) Gram Staining
 
Seminar of U.V. Spectroscopy by SAMIR PANDA
 Seminar of U.V. Spectroscopy by SAMIR PANDA Seminar of U.V. Spectroscopy by SAMIR PANDA
Seminar of U.V. Spectroscopy by SAMIR PANDA
 

Role of AI & ML in beauty care industry

  • 1. Role of AI & ML in Beauty Care Industry GeekNight :: 14th November, 2018 Piyush Bhargava | (Hyderabad)
  • 2. Agenda ● What is AI & ML ? ● AI & ML in Beauty Care ● Examples ● Skin Analysis ● Hair Analysis ● Virtual Makeup ● Live demo of Virtual Makeup ● Q&A 2
  • 3. Artificial Intelligence vs Machine Learning ? ● AI is not a system, but a system has AI ● ML is one of the ways to make a system AI enabled ● ML is making a machine learn ○ from Experience (E) - Training Data ○ w.r.t some Task (T) - Problem Statement ○ and a Performance Measure (P) - Accuracy when P improves with E Artificial Intelligence Machine Learning Deep Learning 3
  • 4. AI & ML in Beauty Care ● Product Recommendation ● Product Personalisation ● Product Validation & Research ● Retail Experiences ○ Virtual (Magic) Mirrors ○ Virtual Try On Apps (Web & Mobile) ● Constant Learning AI Systems 4
  • 5. Some big p:)ayers | Unique So:)utions Recommendation Engines ● Proven - Personalised Skin Care ● My Beauty Matches - Personalised Product Recommendation ● Function of Beauty - Personalised Hair Care ● Beauty.ai - Deep Learning based Beauty Contest ● Olay - Skin analysis from Selfie Augmented Reality ● Modiface - SDKs for Beauty and Face, AR Mirror ● Perfect Corp. - Virtual Makeup Mobile app ● Coty - Magic Mirror for Bourjois Velvet Collection 5
  • 6. Some big p:)ayers | Unique So:)utions 6Video Source: https://www.coty.com/in-the-news/press-release/coty-magic-mirror
  • 7. Product Validation ● Objective - Improve product performance ● Experimental Cosmetics tested on subjects ● Skin / Hair analysis performed through ML / CV modules ● Based on the analysis, ingredients are finalised for the specific product ● At product launch, statistical data used to back the performance claims 7
  • 8. Product Recommendation 8 Quiz / Customer Profiling Matching Product Recommendations A sample product markup table Visitors profile estimation
  • 9. Skin Analysis ● Skin Advisor Applications ● Mobile Skin & Hair Analysis Devices ○ Octa Core Processor ○ 30 to 500 times high magnification lens ○ Normal and UV light sensors ○ Android based, WiFi mirroring ● Pore Analysis ● Moisture / Oiliness ● Wrinkles ● Melanin (Color) / Pigmentation ● Sebum / Acne (T & U Zone) ● Morphological Operations, Edge and Contour analysis, Color based Segmentation, GLCM 9Image Source: http://www.aramhuvis.com/en/apm/
  • 10. Skin Analysis :: Machine Learning Approach 10 Used for detection of symptoms and sometimes even categorisation of symptoms based on severity (high, medium, low) ● Crow’s feet ● Blackheads ● Acne ● Roughness / Dryness ● Eye bags Image Pre-processing RoI Selection Features Extraction Classifier Training Feature Vector Classifier Model Detection / Recognition
  • 11. Skin Color Analysis using ICA Such a tool helps in ● Studying the current pigmentation composition of a subject’s skin ● Recommending specific skin creams based on the analysis above ● Simulate the proposed improvements in skin through reconstruction ● Observe and monitor the actual changes over a period of time 11 Image ICA Haemoglobin (H) Melanin (M) Modified (H’) Modified (M’) Inverse ICA/PCA PCA Reconstructed Image
  • 12. Hair / Scalp Surface Analysis ● Hair Cuticle Damage Analysis ○ Uplifted Cuticle ○ Chipped Cuticle ○ Missing Cuticle ● Broken and Split ends detection ● Hair Density Analysis ● Sebum / Acne analysis on Scalp Surface ● High Magnification Mobile Device used for capturing the Hair / Scalp images for further processing ● The tool helps in assessing the hair and scalp surface, identifying a health index for hair and recommend the most suitable Shampoo and Conditioner (and even hair oil) 12 Image source: https://www.youtube.com/watch?v=FmSkffcqBxY&t=700s Image source: https://www.cgtrader.com/3d-models/science/medical/damaged-hair-cuticle-layer-and-healthy-hair-cuticle-layer
  • 13. Virtual Try On / Makeup ● Simulating makeup on self ● Google Search for “Virtual Makeup” shows many online portals offering such services ● All major cosmetic brands now offer Virtual Makeup feature on their website and mobile apps ○ Try makeup on Eyes, Lips and Face, Hair Color ○ Choose from different shades ○ Add product to cart ○ Order online or collect from nearest store ● Convenient and Comfortable ● More choices than a “brick and mortar” shop ● A better visualization of products on self ● Look Good, Feel Better factor increases popularity among common people ● Steady source of revenue for companies like Modiface, Taaz, Perfect365, etc offering Virtual Makeover Platform / Engine as a product to many cosmetic companies 13 Image Source: http://www.flickr.com/photos/idhren/6236954553/
  • 14. Shape Predictor / Face Landmark Extraction ● Shape predictor is used in analysis of structures and sub regions of an object ● For faces, landmark points describe the shape of the face attributes like: eyes, eyebrows, nose, mouth, jawline & chin ● Wide range of applications, including: face recognition, face filters (snapchat style), face morphing, emotion recognition, blink detection, ... ● Standard pre-trained detectors available ● New shape predictors can be trained ● New dataset can also be generated 68 landmark points generated by Dlib 14
  • 15. Training a Shape Predictor using Dlib ● Create or Pick/Download a large dataset of images ○ Mirroring and Augmentation increase the dataset size further ● Generate landmark annotations (.xml file) ○ Face region :: <top, left, width, height> ○ N points :: <x, y> where N is the number of landmarks we want to train for ● CNN training options affect Size, Speed and Accuracy ● Manually fine-tuned in order to get the desired performance ● Face detection (like OpenCV Haar Cascade, Dlib HOG Detector, CNN detectors) ● The points must be ○ Manually labelled ○ Manually reviewed and corrected ● Example python code (http://dlib.net/train_shape_predictor.py.html) ● Example c++ code (http://dlib.net/train_shape_predictor_ex.cpp.html) ● Also, there is a blog post on Medium that explains training a shape predictor using Dlib 15
  • 16. A generic pipeline for Virtual Makeup 1. Face Detection a. (HAAR, SVM+HOG detector, CNN) 2. Facial Landmark Points extraction a. (using DLib) 3. Lip Mask Creation through Interpolation and Curve Fitting 4. Overlaying Lipstick color through Alpha Compositing 5. Refinements (for more realistic effects) 6. Similar stages for other makeup like eye liner, eye shadow, etc. 16
  • 17. Facial Landmark Points Extraction using DLib ● Deep Learning based implementation ● Can be trained for custom number of landmark points ● Real time, good for Video / Live processing ● Landmark points help in ○ Head Pose estimation, ○ Facial Shape analysis ○ Facial Region Extraction and Analysis ○ Many more Image source: https://ibug.doc.ic.ac.uk/resources/facial-point-annotations/ 17
  • 18. Lip Mask Creation Normal Smiling Open Mouth Lip Contour Lip landmark points 18 Extract lip points Interpolate to get smoother lip contour Create a Mask Blur the Mask 68 landmark points Apply Lipstick
  • 19. Alpha Compositing - Augmenting on Live Frame ● Alpha Blending ensures ○ Smoother edges ○ Lesser Artifacts ○ Transparency effect to retain texture from original image + HSV Colored Lip Region Blurred Mask for Lip Region Without Alpha Blending With Alpha Blending 19
  • 20. Let’s have some fun :) Choose a Lipstick shade Choose a shade for Eyeshadow Virtual Makeup in action ;) 20
  • 21. Cha:((enges & Future Landscape ● Data ● Constant innovation to stay ahead of the Competition ● Voice based systems ● Rendering Realistic Effects ○ Using Skin Surface Reflective Properties ○ Using Ambient Lighting Information ○ Compute intensive ● Shift towards Natural and Organic Products ● Personalised Designing & Packaging of Products ● Shifting lab based cosmetic tests to home environments ○ Using cheaper but reliable sensors ○ Increases the data sources 21
  • 22. Acknowledgements ● Adrian Rosebrock for online tutorials at PyImageSearch ● Intelligent Behavior Understanding Group (iBUG) for providing annotated datasets ● Luca Anzalone for his Medium blog on Shape Predictor Training ● REES46 for their Medium blog post on Progressive Personalization ● ThoughtWorks Hyderabad ● The Awesome Audience .. that’s You !! :) 22
  • 23. Q&A 23 Thank You !! Piyush Bhargava | (Hyderabad) piyush.bhargava@thoughtworks.com