Norbert  Ernser

Norbert Ernser

1641793860

Ce face un blockchain sigur?

Cât de sigură este tehnologia blockchain? Vom explora tehnicile care fac blockchain-ul sigur în continuare în acest articol.

Cum funcționează Blockchain

În anumite privințe, tehnologia blockchain este ca internetul, care se bazează mai degrabă pe o rețea descentralizată decât pe un singur server.

Blockchain folosește un registru descentralizat sau distribuit, care există pe o serie de computere independente, adesea numite noduri, pentru a urmări, a anunța și a coordona tranzacțiile sincronizate. Acest lucru diferă de modelele tradiționale de tranzacționare care se bazează pe o casă de compensare sau bursă care urmărește totul într-un registru central.

Fiecare nod din blockchain-ul descentralizat organizează în mod constant noi date în blocuri și le conectează într-un mod „numai adăugare”. Această structură numai de atașare este o parte importantă a securității blockchain. Nimeni de pe niciun nod nu poate modifica sau șterge datele din blocurile anterioare - pot doar adăuga la lanț. Faptul că lanțul poate fi adăugat numai este una dintre caracteristicile de securitate de bază ale blockchain-ului.

Făcând referire la lanț, participanții pot confirma tranzacțiile. Ea elimină necesitatea unei autorități centrale de compensare.

Ce face un blockchain sigur?

Blockchain-urile sunt securizate printr-o varietate de mecanisme care includ tehnici criptografice avansate și modele matematice de comportament și de luare a deciziilor. Tehnologia Blockchain este structura de bază a majorității sistemelor de criptomonede și este cea care împiedică duplicarea sau distrugerea acestui tip de bani digitali.

Utilizarea tehnologiei blockchain este, de asemenea, explorată în alte contexte în care imuabilitatea și securitatea datelor sunt foarte valoroase. Câteva exemple includ actul de înregistrare și urmărire a donațiilor de caritate, baze de date medicale și managementul lanțului de aprovizionare.

Cu toate acestea, securitatea blockchain este departe de a fi un subiect simplu. Prin urmare, este important să înțelegem conceptele și mecanismele de bază care acordă protecție robustă acestor sisteme inovatoare.

Securitate blockchain: blocuri centralizate

După cum sugerează și numele, „Blockchain Technology” este un blockchain digital centralizat care conține bani pentru tranzacții. Fiecare bloc din acesta este conectat la alte blocuri prezente deasupra și dedesubt. Soluțiile Blockchain sunt special concepute pentru a fi inviolabile și sunt extrem de sigure.

Blockchain definește trei strategii de bază : Criptografie, Descentralizare și Consens.

Toate aceste măsuri fac dificilă modificarea tehnologiei blockchain cu o înregistrare individuală. Acest lucru se datorează faptului că hackerului ar trebui să schimbe întregul bloc care conține înregistrările blockchain. Participanții la rețea au, de asemenea, cheile de securitate atribuite la cheia tranzacțională finală care acționează ca o semnătură digitală personalizată.

În orice circumstanță, dacă înregistrarea este modificată, semnătura digitală devine invalidă și orice atac va fi detectat imediat de rețeaua corespunzătoare.

Într-un blockchain, notificările și alertele timpurii sunt semnificative pentru a preveni alte daune sau atacuri. La o privire, securitatea blockchain oferă, de asemenea, diverse caracteristici de dorit care ajută la securizarea datelor despre tranzacții. Dacă doriți să aflați mai multe despre modul în care blockchain poate schimba viitorul, citiți mai multe despre de ce blockchain este viitorul. 

Caracteristicile cheie ale securității blockchain

Tehnologia Blockchain vine cu multe funcții încorporate care fac apel la obiective precum tranzacțiile cu criptomonede, înregistrările funciare și multe altele. Protecția datelor confidențiale este un factor determinant pentru utilizatori.

Blockchain este, fără îndoială, o măsură care asigură securitatea. A devenit o tehnologie disruptivă datorită capacității sale de a centraliza, stimula, digitiza și securiza validarea tranzacțiilor. Înregistrările sunt protejate prin descentralizarea criptografiei pentru a asigura securitatea blockchain fără greșeală a blocurilor.

Multe sectoare evaluează tehnologia blockchain pentru a detecta diferențierea strategică pentru a-și valorifica potențialul de afaceri. Blockchain are potențialul de a simplifica procesele de afaceri și de a le oferi acces la performanță și securitate îmbunătățite.

Fiecare tranzacție din această tehnologie care este înregistrată și stocată nu este direct etichetată ca blockchain.

Unele dintre caracteristicile definitorii ale tehnologiei blockchain sunt următoarele:

1. Distribuit

Tehnologia blockchain distribuită distribuie uniform procesarea între toți colegii de afaceri, dezvoltând o infrastructură partajată în cadrul sistemului de afaceri. Participanții pot valida în mod independent detaliile fără nicio autoritate centralizată.

2. Imuabil

În acest sens, toate tranzacțiile blockchain din tehnologia bazată pe blockchain sunt imuabile. Chiar și criptarea codului se face pentru tranzacții, acoperind efectiv data, ora și informațiile participanților.

3. Bazat pe consens

Tranzacțiile în tehnologia blockchain se execută doar atunci când părțile prezente în rețea le aprobă în mod anonim. Se poate chiar alege să modifice reglementările bazate pe consens pentru a se potrivi circumstanțelor.

4. Semnătura digitală

Această tehnologie blockchain eficientizează schimbul de valori de translație cu semnături digitale unice care depind de cheile publice. În tehnologia blockchain, codurile cheilor private sunt cunoscute doar de proprietarii de chei pentru a dezvolta dovada proprietății. Este o caracteristică critică care evită orice fraudă în gestionarea înregistrărilor blockchain.

5. Persistent

Tehnologia blockchain persistentă tranzacție invalidă detectată de sistemul bazat pe consens. Nu este ușor să anulați tranzacțiile odată ce acestea sunt integrate în registrul blockchain. Din punct de vedere criptografic, blocurile blockchain sunt create și sigilate în interiorul lanțului, ceea ce face dificil pentru hackeri să editeze sau să ștergă blocurile deja dezvoltate și să le pună în rețea.

Arhitectura de securitate în tehnologia Blockchain

Înregistrările blockchain sunt protejate prin criptografie, în care utilizatorii rețelei au propriile lor chei private și securizate. Această cheie este atribuită direct cheilor de tranzacție și acționează ca o semnătură digitală personalizată. Blockchain protejează și previne orice atac rău intenționat și securizează datele companiei prin criptarea acestora.

Această tehnologie permite întreprinderilor să-și separe responsabilitățile și rolurile. De asemenea, îi ajută să protejeze informațiile confidențiale fără a compromite accesul oricărui utilizator privilegiat.

Blockchain este împuternicit la mai multe straturi pentru dezvoltarea oricărei aplicații de întreprindere.

Aceste straturi includ următoarele elemente:

  • Stratul de acces utilizator
  • Stratul de prezentare
  • Stratul de gestionare a identității și a accesului
  • Strat de aplicație
  • Stratul de rețea
  • Stratul de infrastructură

Tehnologia Blockchain este special concepută pentru a da roade pe termen lung. Profesioniștii în blockchain vă pot ajuta să implementați măsurile esențiale pentru a asigura succesul soluției dvs. blockchain. Blockchain are blocuri de date ireversibile și foarte criptate, care ne permit să luptăm împotriva oricărei conduite frauduloase.

Tehnologia Blockchain protejează, de asemenea, cheile de criptare cu ajutorul standardelor de securitate de înaltă calitate.

Criptoeconomie

Pe lângă criptografie, un concept relativ nou cunoscut sub numele de criptoeconomia joacă, de asemenea, un rol în menținerea securității rețelelor blockchain. Este legat de un domeniu de studiu cunoscut sub numele de teoria jocurilor, care modelează matematic luarea deciziilor de către actori raționali în situații cu reguli și recompense predefinite. În timp ce teoria tradițională a jocurilor poate fi aplicată pe scară largă într-o serie de cazuri, criptoeconomia modelează și descrie în mod specific comportamentul nodurilor pe sistemele blockchain distribuite.

Pe scurt, criptoeconomia este studiul economiei în cadrul protocoalelor blockchain și al posibilelor rezultate pe care proiectarea acestora le poate prezenta pe baza comportamentului participanților săi. Securitatea prin criptoeconomie se bazează pe ideea că sistemele blockchain oferă stimulente mai mari pentru ca nodurile să acționeze cinstit decât să adopte comportamente rău intenționate sau defecte. Încă o dată, algoritmul de consens Proof of Work utilizat în minerit Bitcoin oferă un bun exemplu al acestei structuri de stimulente.

Când Satoshi Nakamoto a creat cadrul pentru minerit Bitcoin, acesta a fost conceput în mod intenționat pentru a fi un proces costisitor și consumatoare de resurse. Datorită complexității și cerințelor sale de calcul, mineritul PoW implică o investiție considerabilă de bani și timp - indiferent de unde și cine este nodul de minerit. Prin urmare, o astfel de structură oferă un puternic descurajare pentru activitățile rău intenționate și stimulente semnificative pentru activitatea minieră cinstită. Nodurile necinstite sau ineficiente vor fi expulzate rapid din rețeaua blockchain, în timp ce minerii cinstiți și eficienți au potențialul de a obține recompense substanțiale în bloc.

În mod similar, acest echilibru de riscuri și recompense oferă, de asemenea, protecție împotriva potențialelor atacuri care ar putea submina consensul, punând rata de hash majoritară a unei rețele blockchain în mâinile unui singur grup sau entitate. Astfel de atacuri, cunoscute sub numele de atacuri de 51%, ar putea fi extrem de dăunătoare dacă sunt executate cu succes. Datorită competitivității minării Proof of Work și amplorii rețelei Bitcoin, probabilitatea ca un actor rău intenționat să câștige controlul asupra majorității nodurilor este extrem de minimă.

În plus, costul puterii de calcul necesar pentru a obține un control de 51% al unei rețele blockchain uriașe ar fi astronomic, oferind un stimulent imediat pentru a face o investiție atât de mare pentru o recompensă potențială relativ mică. Acest fapt contribuie la o caracteristică a blockchain-urilor cunoscută sub numele de Byzantine Fault Tolerance (BFT), care este în esență capacitatea unui sistem distribuit de a continua să funcționeze normal, chiar dacă unele noduri devin compromise sau acționează rău intenționat. 

Atâta timp cât costul înființării majorității nodurilor rău intenționate rămâne prohibitiv și există stimulente mai bune pentru activitatea cinstită, sistemul va putea prospera fără întreruperi semnificative. Cu toate acestea, merită remarcat faptul că rețelele mici de tip blockchain sunt cu siguranță susceptibile de atacuri majoritare, deoarece rata hash totală dedicată acelor sisteme este considerabil mai mică decât cea a Bitcoin.

Gânduri de închidere

Prin utilizarea combinată a teoriei jocurilor și a criptografiei, blockchain-urile sunt capabile să atingă niveluri ridicate de securitate ca sisteme distribuite. La fel ca în aproape toate sistemele, totuși, este esențial ca aceste două domenii de cunoaștere să fie aplicate corect. Un echilibru atent între descentralizare și securitate este vital pentru construirea unei rețele de criptomonede fiabile și eficiente.

Pe măsură ce utilizările blockchain-ului continuă să evolueze, sistemele lor de securitate se vor schimba și ele pentru a răspunde nevoilor diferitelor aplicații. Blockchain-urile private dezvoltate acum pentru întreprinderile de afaceri, de exemplu, se bazează mult mai mult pe securitate prin controlul accesului decât pe mecanismele de teoria jocurilor (sau criptoeconomia) care sunt indispensabile pentru siguranța majorității blockchain-urilor publice.

What is GEEK

Buddha Community

Royce  Reinger

Royce Reinger

1672193100

Face Recognition & Facial Attribute Analysis Library for Python

deepface

Deepface is a lightweight face recognition and facial attribute analysis (age, gender, emotion and race) framework for python. It is a hybrid face recognition framework wrapping state-of-the-art models: VGG-Face, Google FaceNet, OpenFace, Facebook DeepFace, DeepID, ArcFace, Dlib and SFace.

Experiments show that human beings have 97.53% accuracy on facial recognition tasks whereas those models already reached and passed that accuracy level.

Installation  

The easiest way to install deepface is to download it from PyPI. It's going to install the library itself and its prerequisites as well.

$ pip install deepface

DeepFace is also available at Conda. You can alternatively install the package via conda.

$ conda install -c conda-forge deepface

Then you will be able to import the library and use its functionalities.

from deepface import DeepFace

Facial Recognition - Demo

A modern face recognition pipeline consists of 5 common stages: detect, align, normalize, represent and verify. While Deepface handles all these common stages in the background, you don’t need to acquire in-depth knowledge about all the processes behind it. You can just call its verification, find or analysis function with a single line of code.

Face Verification - Demo

This function verifies face pairs as same person or different persons. It expects exact image paths as inputs. Passing numpy or base64 encoded images is also welcome. Then, it is going to return a dictionary and you should check just its verified key.

result = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg")

Face recognition - Demo

Face recognition requires applying face verification many times. Herein, deepface has an out-of-the-box find function to handle this action. It's going to look for the identity of input image in the database path and it will return pandas data frame as output.

df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db")

Embeddings

Face recognition models basically represent facial images as multi-dimensional vectors. Sometimes, you need those embedding vectors directly. DeepFace comes with a dedicated representation function.

embedding = DeepFace.represent(img_path = "img.jpg")

This function returns an array as output. The size of the output array would be different based on the model name. For instance, VGG-Face is the default model for deepface and it represents facial images as 2622 dimensional vectors.

assert isinstance(embedding, list)
assert model_name = "VGG-Face" and len(embedding) == 2622

Here, embedding is also plotted with 2622 slots horizontally. Each slot is corresponding to a dimension value in the embedding vector and dimension value is explained in the colorbar on the right. Similar to 2D barcodes, vertical dimension stores no information in the illustration.

Face recognition models - Demo

Deepface is a hybrid face recognition package. It currently wraps many state-of-the-art face recognition models: VGG-Face , Google FaceNet, OpenFace, Facebook DeepFace, DeepID, ArcFace, Dlib and SFace. The default configuration uses VGG-Face model.

models = [
  "VGG-Face", 
  "Facenet", 
  "Facenet512", 
  "OpenFace", 
  "DeepFace", 
  "DeepID", 
  "ArcFace", 
  "Dlib", 
  "SFace",
]

#face verification
result = DeepFace.verify(img1_path = "img1.jpg", 
      img2_path = "img2.jpg", 
      model_name = models[1]
)

#face recognition
df = DeepFace.find(img_path = "img1.jpg",
      db_path = "C:/workspace/my_db", 
      model_name = models[1]
)

#embeddings
embedding = DeepFace.represent(img_path = "img.jpg", 
      model_name = models[1]
)

FaceNet, VGG-Face, ArcFace and Dlib are overperforming ones based on experiments. You can find out the scores of those models below on both Labeled Faces in the Wild and YouTube Faces in the Wild data sets declared by its creators.

ModelLFW ScoreYTF Score
Facenet51299.65%-
SFace99.60%-
ArcFace99.41%-
Dlib99.38 %-
Facenet99.20%-
VGG-Face98.78%97.40%
Human-beings97.53%-
OpenFace93.80%-
DeepID-97.05%

Similarity

Face recognition models are regular convolutional neural networks and they are responsible to represent faces as vectors. We expect that a face pair of same person should be more similar than a face pair of different persons.

Similarity could be calculated by different metrics such as Cosine Similarity, Euclidean Distance and L2 form. The default configuration uses cosine similarity.

metrics = ["cosine", "euclidean", "euclidean_l2"]

#face verification
result = DeepFace.verify(img1_path = "img1.jpg", 
          img2_path = "img2.jpg", 
          distance_metric = metrics[1]
)

#face recognition
df = DeepFace.find(img_path = "img1.jpg", 
          db_path = "C:/workspace/my_db", 
          distance_metric = metrics[1]
)

Euclidean L2 form seems to be more stable than cosine and regular Euclidean distance based on experiments.

Facial Attribute Analysis - Demo

Deepface also comes with a strong facial attribute analysis module including age, gender, facial expression (including angry, fear, neutral, sad, disgust, happy and surprise) and race (including asian, white, middle eastern, indian, latino and black) predictions.

obj = DeepFace.analyze(img_path = "img4.jpg", 
        actions = ['age', 'gender', 'race', 'emotion']
)

Age model got ± 4.65 MAE; gender model got 97.44% accuracy, 96.29% precision and 95.05% recall as mentioned in its tutorial.

Face Detectors - Demo

Face detection and alignment are important early stages of a modern face recognition pipeline. Experiments show that just alignment increases the face recognition accuracy almost 1%. OpenCV, SSD, Dlib, MTCNN, RetinaFace and MediaPipe detectors are wrapped in deepface.

All deepface functions accept an optional detector backend input argument. You can switch among those detectors with this argument. OpenCV is the default detector.

backends = [
  'opencv', 
  'ssd', 
  'dlib', 
  'mtcnn', 
  'retinaface', 
  'mediapipe'
]

#face verification
obj = DeepFace.verify(img1_path = "img1.jpg", 
        img2_path = "img2.jpg", 
        detector_backend = backends[4]
)

#face recognition
df = DeepFace.find(img_path = "img.jpg", 
        db_path = "my_db", 
        detector_backend = backends[4]
)

#embeddings
embedding = DeepFace.represent(img_path = "img.jpg", 
        detector_backend = backends[4]
)

#facial analysis
demography = DeepFace.analyze(img_path = "img4.jpg", 
        detector_backend = backends[4]
)

#face detection and alignment
face = DeepFace.detectFace(img_path = "img.jpg", 
        target_size = (224, 224), 
        detector_backend = backends[4]
)

Face recognition models are actually CNN models and they expect standard sized inputs. So, resizing is required before representation. To avoid deformation, deepface adds black padding pixels according to the target size argument after detection and alignment.

RetinaFace and MTCNN seem to overperform in detection and alignment stages but they are much slower. If the speed of your pipeline is more important, then you should use opencv or ssd. On the other hand, if you consider the accuracy, then you should use retinaface or mtcnn.

The performance of RetinaFace is very satisfactory even in the crowd as seen in the following illustration. Besides, it comes with an incredible facial landmark detection performance. Highlighted red points show some facial landmarks such as eyes, nose and mouth. That's why, alignment score of RetinaFace is high as well.

You can find out more about RetinaFace on this repo.

Real Time Analysis - Demo

You can run deepface for real time videos as well. Stream function will access your webcam and apply both face recognition and facial attribute analysis. The function starts to analyze a frame if it can focus a face sequentially 5 frames. Then, it shows results 5 seconds.

DeepFace.stream(db_path = "C:/User/Sefik/Desktop/database")

Even though face recognition is based on one-shot learning, you can use multiple face pictures of a person as well. You should rearrange your directory structure as illustrated below.

user
├── database
│   ├── Alice
│   │   ├── Alice1.jpg
│   │   ├── Alice2.jpg
│   ├── Bob
│   │   ├── Bob.jpg

API - Demo

Deepface serves an API as well. You can clone /api/api.py and pass it to python command as an argument. This will get a rest service up. In this way, you can call deepface from an external system such as mobile app or web.

python api.py

Face recognition, facial attribute analysis and vector representation functions are covered in the API. You are expected to call these functions as http post methods. Service endpoints will be http://127.0.0.1:5000/verify for face recognition, http://127.0.0.1:5000/analyze for facial attribute analysis, and http://127.0.0.1:5000/represent for vector representation. You should pass input images as base64 encoded string in this case. Here, you can find a postman project.

Command Line Interface

DeepFace comes with a command line interface as well. You are able to access its functions in command line as shown below. The command deepface expects the function name as 1st argument and function arguments thereafter.

#face verification
$ deepface verify -img1_path tests/dataset/img1.jpg -img2_path tests/dataset/img2.jpg

#facial analysis
$ deepface analyze -img_path tests/dataset/img1.jpg

Tech Stack - Vlog, Tutorial

Face recognition models represent facial images as vector embeddings. The idea behind facial recognition is that vectors should be more similar for same person than different persons. The question is that where and how to store facial embeddings in a large scale system. Tech stack is vast to store vector embeddings. To determine the right tool, you should consider your task such as face verification or face recognition, priority such as speed or confidence, and also data size.

Contribution 

Pull requests are welcome! You should run the unit tests locally by running test/unit_tests.py. Once a PR sent, GitHub test workflow will be run automatically and unit test results will be available in GitHub actions before approval.

Support

There are many ways to support a project - starring⭐️ the GitHub repo is just one 🙏

You can also support this work on Patreon

 

Citation

Please cite deepface in your publications if it helps your research. Here are its BibTex entries:

If you use deepface for facial recogntion purposes, please cite the this publication.

@inproceedings{serengil2020lightface,
  title        = {LightFace: A Hybrid Deep Face Recognition Framework},
  author       = {Serengil, Sefik Ilkin and Ozpinar, Alper},
  booktitle    = {2020 Innovations in Intelligent Systems and Applications Conference (ASYU)},
  pages        = {23-27},
  year         = {2020},
  doi          = {10.1109/ASYU50717.2020.9259802},
  url          = {https://doi.org/10.1109/ASYU50717.2020.9259802},
  organization = {IEEE}
}

If you use deepface for facial attribute analysis purposes such as age, gender, emotion or ethnicity prediction, please cite the this publication.

@inproceedings{serengil2021lightface,
  title        = {HyperExtended LightFace: A Facial Attribute Analysis Framework},
  author       = {Serengil, Sefik Ilkin and Ozpinar, Alper},
  booktitle    = {2021 International Conference on Engineering and Emerging Technologies (ICEET)},
  pages        = {1-4},
  year         = {2021},
  doi          = {10.1109/ICEET53442.2021.9659697},
  url          = {https://doi.org/10.1109/ICEET53442.2021.9659697},
  organization = {IEEE}
}

Also, if you use deepface in your GitHub projects, please add deepface in the requirements.txt.

Download Details:

Author: Serengil
Source Code: https://github.com/serengil/deepface 
License: MIT license

#machinelearning #python #deeplearning 

A Lightweight Face Recognition and Facial Attribute Analysis

deepface

Deepface is a lightweight face recognition and facial attribute analysis (age, gender, emotion and race) framework for python. It is a hybrid face recognition framework wrapping state-of-the-art models: VGG-Face, Google FaceNet, OpenFace, Facebook DeepFace, DeepID, ArcFace and Dlib.

Experiments show that human beings have 97.53% accuracy on facial recognition tasks whereas those models already reached and passed that accuracy level.

Installation

The easiest way to install deepface is to download it from PyPI. It's going to install the library itself and its prerequisites as well. The library is mainly based on TensorFlow and Keras.

pip install deepface

Then you will be able to import the library and use its functionalities.

from deepface import DeepFace

Facial Recognition - Demo

A modern face recognition pipeline consists of 5 common stages: detect, align, normalize, represent and verify. While Deepface handles all these common stages in the background, you don’t need to acquire in-depth knowledge about all the processes behind it. You can just call its verification, find or analysis function with a single line of code.

Face Verification - Demo

This function verifies face pairs as same person or different persons. It expects exact image paths as inputs. Passing numpy or based64 encoded images is also welcome. Then, it is going to return a dictionary and you should check just its verified key.

result = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg")

Face recognition - Demo

Face recognition requires applying face verification many times. Herein, deepface has an out-of-the-box find function to handle this action. It's going to look for the identity of input image in the database path and it will return pandas data frame as output.

df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db")

Face recognition models - Demo

Deepface is a hybrid face recognition package. It currently wraps many state-of-the-art face recognition models: VGG-Face , Google FaceNet, OpenFace, Facebook DeepFace, DeepID, ArcFace and Dlib. The default configuration uses VGG-Face model.

models = ["VGG-Face", "Facenet", "Facenet512", "OpenFace", "DeepFace", "DeepID", "ArcFace", "Dlib"]

#face verification
result = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg", model_name = models[1])

#face recognition
df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db", model_name = models[1])

FaceNet, VGG-Face, ArcFace and Dlib are overperforming ones based on experiments. You can find out the scores of those models below on both Labeled Faces in the Wild and YouTube Faces in the Wild data sets declared by its creators.

ModelLFW ScoreYTF Score
Facenet51299.65%-
ArcFace99.41%-
Dlib99.38 %-
Facenet99.20%-
VGG-Face98.78%97.40%
Human-beings97.53%-
OpenFace93.80%-
DeepID-97.05%

Similarity

Face recognition models are regular convolutional neural networks and they are responsible to represent faces as vectors. We expect that a face pair of same person should be more similar than a face pair of different persons.

Similarity could be calculated by different metrics such as Cosine Similarity, Euclidean Distance and L2 form. The default configuration uses cosine similarity.

metrics = ["cosine", "euclidean", "euclidean_l2"]

#face verification
result = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg", distance_metric = metrics[1])

#face recognition
df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db", distance_metric = metrics[1])

Euclidean L2 form seems to be more stable than cosine and regular Euclidean distance based on experiments.

Facial Attribute Analysis - Demo

Deepface also comes with a strong facial attribute analysis module including age, gender, facial expression (including angry, fear, neutral, sad, disgust, happy and surprise) and race (including asian, white, middle eastern, indian, latino and black) predictions.

obj = DeepFace.analyze(img_path = "img4.jpg", actions = ['age', 'gender', 'race', 'emotion'])

Age model got ± 4.65 MAE; gender model got 97.44% accuracy, 96.29% precision and 95.05% recall as mentioned in its tutorial.

Streaming and Real Time Analysis - Demo

You can run deepface for real time videos as well. Stream function will access your webcam and apply both face recognition and facial attribute analysis. The function starts to analyze a frame if it can focus a face sequantially 5 frames. Then, it shows results 5 seconds.

DeepFace.stream(db_path = "C:/User/Sefik/Desktop/database")

Even though face recognition is based on one-shot learning, you can use multiple face pictures of a person as well. You should rearrange your directory structure as illustrated below.

user
├── database
│   ├── Alice
│   │   ├── Alice1.jpg
│   │   ├── Alice2.jpg
│   ├── Bob
│   │   ├── Bob.jpg

Face Detectors - Demo

Face detection and alignment are important early stages of a modern face recognition pipeline. Experiments show that just alignment increases the face recognition accuracy almost 1%. OpenCV, SSD, Dlib, MTCNN and RetinaFace detectors are wrapped in deepface.

All deepface functions accept an optional detector backend input argument. You can switch among those detectors with this argument. OpenCV is the default detector.

backends = ['opencv', 'ssd', 'dlib', 'mtcnn', 'retinaface']

#face verification
obj = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg", detector_backend = backends[4])

#face recognition
df = DeepFace.find(img_path = "img.jpg", db_path = "my_db", detector_backend = backends[4])

#facial analysis
demography = DeepFace.analyze(img_path = "img4.jpg", detector_backend = backends[4])

#face detection and alignment
face = DeepFace.detectFace(img_path = "img.jpg", target_size = (224, 224), detector_backend = backends[4])

Face recognition models are actually CNN models and they expect standard sized inputs. So, resizing is required before representation. To avoid deformation, deepface adds black padding pixels according to the target size argument after detection and alignment.

RetinaFace and MTCNN seem to overperform in detection and alignment stages but they are much slower. If the speed of your pipeline is more important, then you should use opencv or ssd. On the other hand, if you consider the accuracy, then you should use retinaface or mtcnn.

The performance of RetinaFace is very satisfactory even in the crowd as seen in the following illustration. Besides, it comes with an incredible facial landmark detection performance. Highlighted red points show some facial landmarks such as eyes, nose and mouth. That's why, alignment score of RetinaFace is high as well.

You can find out more about RetinaFace on this repo.

API - Demo

Deepface serves an API as well. You can clone /api/api.py and pass it to python command as an argument. This will get a rest service up. In this way, you can call deepface from an external system such as mobile app or web.

python api.py

Face recognition, facial attribute analysis and vector representation functions are covered in the API. You are expected to call these functions as http post methods. Service endpoints will be http://127.0.0.1:5000/verify for face recognition, http://127.0.0.1:5000/analyze for facial attribute analysis, and http://127.0.0.1:5000/represent for vector representation. You should pass input images as base64 encoded string in this case. Here, you can find a postman project.

Tech Stack - Vlog, Tutorial

Face recognition models represent facial images as vector embeddings. The idea behind facial recognition is that vectors should be more similar for same person than different persons. The question is that where and how to store facial embeddings in a large scale system. Herein, deepface offers a represention function to find vector embeddings from facial images.

embedding = DeepFace.represent(img_path = "img.jpg", model_name = 'Facenet')

Tech stack is vast to store vector embeddings. To determine the right tool, you should consider your task such as face verification or face recognition, priority such as speed or confidence, and also data size.

Contribution

Pull requests are welcome. You should run the unit tests locally by running test/unit_tests.py. Please share the unit test result logs in the PR. Deepface is currently compatible with TF 1 and 2 versions. Change requests should satisfy those requirements both.

Support

There are many ways to support a project - starring⭐️ the GitHub repo is just one 🙏

You can also support this work on Patreon

 

Citation

Please cite deepface in your publications if it helps your research. Here are its BibTeX entries:

@inproceedings{serengil2020lightface,
  title        = {LightFace: A Hybrid Deep Face Recognition Framework},
  author       = {Serengil, Sefik Ilkin and Ozpinar, Alper},
  booktitle    = {2020 Innovations in Intelligent Systems and Applications Conference (ASYU)},
  pages        = {23-27},
  year         = {2020},
  doi          = {10.1109/ASYU50717.2020.9259802},
  url          = {https://doi.org/10.1109/ASYU50717.2020.9259802},
  organization = {IEEE}
}
@inproceedings{serengil2021lightface,
  title        = {HyperExtended LightFace: A Facial Attribute Analysis Framework},
  author       = {Serengil, Sefik Ilkin and Ozpinar, Alper},
  booktitle    = {2021 International Conference on Engineering and Emerging Technologies (ICEET)},
  pages        = {1-4},
  year         = {2021},
  doi          = {10.1109/ICEET53442.2021.9659697},
  url.         = {https://doi.org/10.1109/ICEET53442.2021.9659697},
  organization = {IEEE}
}

Also, if you use deepface in your GitHub projects, please add deepface in the requirements.txt.

Author: Serengil
Source Code: https://github.com/serengil/deepface 
License: MIT License

#python #machine-learning 

Dominic  Feeney

Dominic Feeney

1648217849

Deepface: A Face Recognition and Facial Attribute Analysis for Python

deepface

Deepface is a lightweight face recognition and facial attribute analysis (age, gender, emotion and race) framework for python. It is a hybrid face recognition framework wrapping state-of-the-art models: VGG-Face, Google FaceNet, OpenFace, Facebook DeepFace, DeepID, ArcFace and Dlib.

Experiments show that human beings have 97.53% accuracy on facial recognition tasks whereas those models already reached and passed that accuracy level.

Installation

The easiest way to install deepface is to download it from PyPI. It's going to install the library itself and its prerequisites as well. The library is mainly powered by TensorFlow and Keras.

pip install deepface

Then you will be able to import the library and use its functionalities.

from deepface import DeepFace

Facial Recognition - Demo

A modern face recognition pipeline consists of 5 common stages: detect, align, normalize, represent and verify. While Deepface handles all these common stages in the background, you don’t need to acquire in-depth knowledge about all the processes behind it. You can just call its verification, find or analysis function with a single line of code.

Face Verification - Demo

This function verifies face pairs as same person or different persons. It expects exact image paths as inputs. Passing numpy or based64 encoded images is also welcome. Then, it is going to return a dictionary and you should check just its verified key.

result = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg")

Face recognition - Demo

Face recognition requires applying face verification many times. Herein, deepface has an out-of-the-box find function to handle this action. It's going to look for the identity of input image in the database path and it will return pandas data frame as output.

df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db")

Face recognition models - Demo

Deepface is a hybrid face recognition package. It currently wraps many state-of-the-art face recognition models: VGG-Face , Google FaceNet, OpenFace, Facebook DeepFace, DeepID, ArcFace and Dlib. The default configuration uses VGG-Face model.

models = ["VGG-Face", "Facenet", "Facenet512", "OpenFace", "DeepFace", "DeepID", "ArcFace", "Dlib"]

#face verification
result = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg", model_name = models[1])

#face recognition
df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db", model_name = models[1])

FaceNet, VGG-Face, ArcFace and Dlib are overperforming ones based on experiments. You can find out the scores of those models below on both Labeled Faces in the Wild and YouTube Faces in the Wild data sets declared by its creators.

ModelLFW ScoreYTF Score
Facenet51299.65%-
ArcFace99.41%-
Dlib99.38 %-
Facenet99.20%-
VGG-Face98.78%97.40%
Human-beings97.53%-
OpenFace93.80%-
DeepID-97.05%

Similarity

Face recognition models are regular convolutional neural networks and they are responsible to represent faces as vectors. We expect that a face pair of same person should be more similar than a face pair of different persons.

Similarity could be calculated by different metrics such as Cosine Similarity, Euclidean Distance and L2 form. The default configuration uses cosine similarity.

metrics = ["cosine", "euclidean", "euclidean_l2"]

#face verification
result = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg", distance_metric = metrics[1])

#face recognition
df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db", distance_metric = metrics[1])

Euclidean L2 form seems to be more stable than cosine and regular Euclidean distance based on experiments.

Facial Attribute Analysis - Demo

Deepface also comes with a strong facial attribute analysis module including age, gender, facial expression (including angry, fear, neutral, sad, disgust, happy and surprise) and race (including asian, white, middle eastern, indian, latino and black) predictions.

obj = DeepFace.analyze(img_path = "img4.jpg", actions = ['age', 'gender', 'race', 'emotion'])

Age model got ± 4.65 MAE; gender model got 97.44% accuracy, 96.29% precision and 95.05% recall as mentioned in its tutorial.

Streaming and Real Time Analysis - Demo

You can run deepface for real time videos as well. Stream function will access your webcam and apply both face recognition and facial attribute analysis. The function starts to analyze a frame if it can focus a face sequantially 5 frames. Then, it shows results 5 seconds.

DeepFace.stream(db_path = "C:/User/Sefik/Desktop/database")

Even though face recognition is based on one-shot learning, you can use multiple face pictures of a person as well. You should rearrange your directory structure as illustrated below.

user
├── database
│   ├── Alice
│   │   ├── Alice1.jpg
│   │   ├── Alice2.jpg
│   ├── Bob
│   │   ├── Bob.jpg

Face Detectors - Demo

Face detection and alignment are important early stages of a modern face recognition pipeline. Experiments show that just alignment increases the face recognition accuracy almost 1%. OpenCV, SSD, Dlib, MTCNN, RetinaFace and MediaPipe detectors are wrapped in deepface.

All deepface functions accept an optional detector backend input argument. You can switch among those detectors with this argument. OpenCV is the default detector.

backends = ['opencv', 'ssd', 'dlib', 'mtcnn', 'retinaface', 'mediapipe']

#face verification
obj = DeepFace.verify(img1_path = "img1.jpg", img2_path = "img2.jpg", detector_backend = backends[4])

#face recognition
df = DeepFace.find(img_path = "img.jpg", db_path = "my_db", detector_backend = backends[4])

#facial analysis
demography = DeepFace.analyze(img_path = "img4.jpg", detector_backend = backends[4])

#face detection and alignment
face = DeepFace.detectFace(img_path = "img.jpg", target_size = (224, 224), detector_backend = backends[4])

Face recognition models are actually CNN models and they expect standard sized inputs. So, resizing is required before representation. To avoid deformation, deepface adds black padding pixels according to the target size argument after detection and alignment.

RetinaFace and MTCNN seem to overperform in detection and alignment stages but they are much slower. If the speed of your pipeline is more important, then you should use opencv or ssd. On the other hand, if you consider the accuracy, then you should use retinaface or mtcnn.

The performance of RetinaFace is very satisfactory even in the crowd as seen in the following illustration. Besides, it comes with an incredible facial landmark detection performance. Highlighted red points show some facial landmarks such as eyes, nose and mouth. That's why, alignment score of RetinaFace is high as well.

You can find out more about RetinaFace on this repo.

API - Demo

Deepface serves an API as well. You can clone /api/api.py and pass it to python command as an argument. This will get a rest service up. In this way, you can call deepface from an external system such as mobile app or web.

python api.py

Face recognition, facial attribute analysis and vector representation functions are covered in the API. You are expected to call these functions as http post methods. Service endpoints will be http://127.0.0.1:5000/verify for face recognition, http://127.0.0.1:5000/analyze for facial attribute analysis, and http://127.0.0.1:5000/represent for vector representation. You should pass input images as base64 encoded string in this case. Here, you can find a postman project.

Tech Stack - Vlog, Tutorial

Face recognition models represent facial images as vector embeddings. The idea behind facial recognition is that vectors should be more similar for same person than different persons. The question is that where and how to store facial embeddings in a large scale system. Herein, deepface offers a represention function to find vector embeddings from facial images.

embedding = DeepFace.represent(img_path = "img.jpg", model_name = 'Facenet')

Tech stack is vast to store vector embeddings. To determine the right tool, you should consider your task such as face verification or face recognition, priority such as speed or confidence, and also data size.

Contribution

Pull requests are welcome. You should run the unit tests locally by running test/unit_tests.py. Please share the unit test result logs in the PR. Deepface is currently compatible with TF 1 and 2 versions. Change requests should satisfy those requirements both.

Support

There are many ways to support a project - starring⭐️ the GitHub repo is just one 🙏

You can also support this work on Patreon

 

Citation

Please cite deepface in your publications if it helps your research. Here are BibTeX entries:

@inproceedings{serengil2020lightface,
  title        = {LightFace: A Hybrid Deep Face Recognition Framework},
  author       = {Serengil, Sefik Ilkin and Ozpinar, Alper},
  booktitle    = {2020 Innovations in Intelligent Systems and Applications Conference (ASYU)},
  pages        = {23-27},
  year         = {2020},
  doi          = {10.1109/ASYU50717.2020.9259802},
  url          = {https://doi.org/10.1109/ASYU50717.2020.9259802},
  organization = {IEEE}
}
@inproceedings{serengil2021lightface,
  title        = {HyperExtended LightFace: A Facial Attribute Analysis Framework},
  author       = {Serengil, Sefik Ilkin and Ozpinar, Alper},
  booktitle    = {2021 International Conference on Engineering and Emerging Technologies (ICEET)},
  pages        = {1-4},
  year         = {2021},
  doi          = {10.1109/ICEET53442.2021.9659697},
  url          = {https://doi.org/10.1109/ICEET53442.2021.9659697},
  organization = {IEEE}
}

Also, if you use deepface in your GitHub projects, please add deepface in the requirements.txt.

Download Details:
Author: serengil
Source Code: https://github.com/serengil/deepface
License: MIT License

#tensorflow  #python #machinelearning 

Devin Pinto

1606217442

Blockchain Certification | Blockchain Training Course | Blockchain Council

In all the market sectors, Blockchain technology has contributed to the redesign. The improvements that were once impossible have been pushed forward. Blockchain is one of the leading innovations with the ability to influence the various sectors of the industry. It also has the ability to be one of the career-influencing innovations at the same time. We have seen an increasing inclination towards the certification of the Blockchain in recent years, and there are obvious reasons behind it. Blockchain has everything to offer, from good packages to its universal application and futuristic development. Let’s address the reasons why one should go for Blockchain certification.

5 advantages of certification by Blockchain:

1. Lucrative packages- Everyone who completes their education or upskills themselves wants to end up with a good bundle, not only is one assured of a good learning experience with Blockchain, but the packages are drool-worthy at the same time. A Blockchain developer’s average salary varies between $150,000 and $175,000 per annum. Comparatively, a software developer gets a $137,000 per year salary. For a Blockchain developer, the San Francisco Bay area provides the highest bundle, amounting to $162,288 per annum. There’s no point arguing that learning about Blockchain is a smart decision with such lucrative packages.

2. Growing industry- When you select any qualification course, it becomes important that you choose a growing segment or industry that promises potential in the future. You should anticipate all of these with Blockchain. The size of the blockchain market is expected to rise from USD 3.0 billion in 2020 to USD 39.7 billion by 2025. This will see an incredible 67.3 percent CAGR between 2020-2025. To help business processes, several businesses are outsourcing Blockchain technologies. This clearly demonstrates that there will be higher demand in the future for Blockchain developers and certified Blockchain professionals.

3. Universal application- One of the major reasons for the success of Blockchain is that it has a global application. It is not sector-specific. Blockchain usage cases are discovered by almost all market segments. In addition, other innovations such as AI, big data, data science and much more are also supported by Blockchain. It becomes easier to get into a suitable industry once you know about Blockchain.

**4. Work protection-**Surely you would like to invest in an ability that ensures job security. You had the same chance for Blockchain. Since this is the technology of the future, understanding that Blockchain can keep up with futuristic developments will help in a successful and safe job.

**5.**After a certain point of your professional life, you are expected to learn about new abilities that can help enhance your skills. Upskilling is paramount. Upskilling oneself has become the need for the hour, and choosing a path that holds a lot of potential for the future is the best way to do this. For all computer geeks and others who want to gain awareness of emerging technology, Blockchain is a good option.

Concluding thoughts- opting for Blockchain certification is a successful career move with all these advantages. You will be able to find yourself in a safe and secured work profile once you have all the knowledge and information. Link for Blockchain certification programme with the Blockchain Council.

#blockchain certificate #blockchain training #blockchain certification #blockchain developers #blockchain #blockchain council

5 Blockchain Applications That Have Transformed the World of Technology

The blockchain is the decentralized database of the blocks of information, which gets recorded in the chain format and linked in a secured crypto graphical manner. This technology ensures proper safety of the data due to its secure nature, and it totally changes how people carry out transactions. It also brings about a faster and secure process of validating information needed to establish reliability.

Though blockchain technology came into the market to carry out only digital transactions, it is now used in various industries like supply chain, finance, health care, and many more.

The blockchain technology has made its position in mobile app development as well. Blockchain applications are transparent and accountable. From getting easy access to medical records and buying insurance, you can see blockchain applications everywhere.

Here are some of the areas where you can see the use of blockchain applications and how they have changed various industries.

1. Ripple

Ripple is useful for increasing banking transactions. The implementation of blockchain technology in the financial sector is much more profound than any other sector. Ripple proves this. It is one of the greatest tools to record and complete financial transactions.

It develops a large network despite strict physical boundaries. As there is no such third-party involvement present, the cost of these transactions is lower than usual. At the same time, the network also remains transparent and quite secured.

It is normally seen that financial transactions that happen globally are

error-prone and take a lot of time. In addition to this, when the transaction

fees and exchange rates get added up, the total cost usually gets high.

However, Ripple offers real-time international transactions without spending too much money. It has the network of about 200+ institutions making the process affordable, secure, and fast for all sorts of international transactions.

2. Etherisc

This blockchain application helps in automating flight insurance. Insurance is another area where blockchain is gaining popularity. Through this application, insurers can make smart contracts rather than getting involved in the traditional contracts that are usually complex. Etherisc is the blockchain application that helps customers buy flight insurance. If the flight gets canceled or delayed, they do not have to wait for months to get the payment back. This application ensures an on-time payout.

#blockchain #blockchain-technology #blockchain-development #blockchain-use-cases #blockchain-a #blockchain-technologies #technology #decentralization