Unlocking Any Door In The 21st Century. Immersion In Biometric Security.

Payment Village
Payment VillageSecurity research lead em Payment Village
Unlocking any door in the 21st century
Immersion in biometric security
1
Timur Yunusov & Alexandra Murzina
Who we are
● ex A-Team Cyber R&D Lab
● Head of research
● Senior ML security expert
2
Outline
● Current state of AI/ML in biometrics
● ML attacks landscape
● Attacking devices
○ Device 1 - undisclosed
○ Device 2 - ZKTeco
○ Device 3 - Eufy
● Conclusions
● Security Checklist
3
United States:
State-specific biometric laws, e.g., BIPA in Illinois and CCPA in California.
FBI uses biometrics for law enforcement and border control.
China:
Extensive government use of biometrics for surveillance and security.
Requirement to store critical data, including biometrics, within the
country.
India:
Aadhaar Act regulates biometric data collected under Aadhaar
program.
Proposed Data Privacy Bill aims for comprehensive data protection.
European Union (EU):
GDPR regulates biometric data with explicit consent and stringent
protection.
United Arab Emirates (UAE):
DIFC's data protection law covers biometric data.
Government uses biometrics extensively for security and services.
Japan:
APPI regulates personal data, including biometrics, with consent and
protection.
Legislation
United Kingdom:
Data Protection Act regulates personal data processing, including
biometrics.
Independent oversight of law enforcement biometric use by Biometrics
Commissioner.
South Korea:
PIPA considers biometric data "sensitive," requiring consent and
protection.
Regulations allow biometric authentication in financial transactions.
Brazil:
LGPD regulates personal data processing, including biometrics, with
consent and protection.
Requires security measures and impact assessments.
South Africa:
POPIA regulates personal data processing, including biometrics, with
consent and protection.
Russia:
Personal Data Law mandates consent for biometric processing.
Federal Law regulates fingerprinting.
Unified Biometric System enables bank identification.
Government uses biometrics for security and law enforcement.
4
Practical aspects
5
Current state of
AI/ML in biometrics
6
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
7
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Early Methods, Eigenfaces
initially, manual analysis of facial features in photos measured distances and angles between
landmarks like eyes and nose. Automated face recognition began in the late 1980s with Eigenfaces,
using PCA to extract features from grayscale images, representing faces as weighted "eigenfaces."
Local Feature Methods
techniques like LBP and Gabor wavelets focused on specific face regions, capturing texture and
local changes.
2D and 3D Face Models
2D and 3D face models accounted for pose and expression variations, with 3D models providing
depth information.
8
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Early Methods, Eigenfaces
initially, manual analysis of facial features in photos measured distances and angles between
landmarks like eyes and nose. Automated face recognition began in the late 1980s with Eigenfaces,
using PCA to extract features from grayscale images, representing faces as weighted "eigenfaces."
Local Feature Methods
techniques like LBP and Gabor wavelets focused on specific face regions, capturing texture and
local changes.
2D and 3D Face Models
2D and 3D face models accounted for pose and expression variations, with 3D models providing
depth information.
Machine Learning and Deep Learning
machine learning and deep learning techniques, like SVMs and CNNs, automatically learned and
extracted facial features from large datasets, enhancing recognition accuracy and robustness.
9
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Early Methods, Eigenfaces
initially, manual analysis of facial features in photos measured distances and angles between
landmarks like eyes and nose. Automated face recognition began in the late 1980s with Eigenfaces,
using PCA to extract features from grayscale images, representing faces as weighted "eigenfaces."
Local Feature Methods
techniques like LBP and Gabor wavelets focused on specific face regions, capturing texture and
local changes.
2D and 3D Face Models
2D and 3D face models accounted for pose and expression variations, with 3D models providing
depth information.
Machine Learning and Deep Learning
machine learning and deep learning techniques, like SVMs and CNNs, automatically learned and
extracted facial features from large datasets, enhancing recognition accuracy and robustness.
Depth Sensing and Infrared Cameras
Modern systems use depth sensing and infrared cameras to capture facial information in
challenging lighting or obscured faces, enabling accurate recognition and spoof detection.
Multi-modal and Fusion Methods
Combining multiple biometric modalities, such as face and voice or fusing 2D and 3D data, has
enhanced recognition performance.
Emotion Recognition and Liveness Detection
Recent advancements include emotion recognition from facial expressions and liveness detection
to verify the subject's presence.
10
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Face Detection
algorithms like Haar cascades or SSD locate and isolate faces in
images or video streams.
Face Alignment
detected faces are transformed into a standard format by
rotating, scaling, and translating them for uniformity.
Feature Extraction
machine learning models, such as CNNs, extract unique facial
features and create a face embedding or feature vector.
Face Matching
extracted features are compared with stored feature vectors
using distance metrics like Euclidean or cosine distance.
Systems identify the closest match or verify if the face matches
a specific representation.
Decision Making
the system determines whether to accept or reject
identification or verification based on matching results,
sometimes providing a confidence score or probability.
11
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Face Detection
algorithms like Haar cascades or SSD locate and isolate faces in
images or video streams.
Face Alignment
detected faces are transformed into a standard format by
rotating, scaling, and translating them for uniformity.
Feature Extraction
machine learning models, such as CNNs, extract unique facial
features and create a face embedding or feature vector.
Face Matching
extracted features are compared with stored feature vectors
using distance metrics like Euclidean or cosine distance.
Systems identify the closest match or verify if the face matches
a specific representation.
Decision Making
the system determines whether to accept or reject
identification or verification based on matching results,
sometimes providing a confidence score or probability.
12
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Face Detection
algorithms like Haar cascades or SSD locate and isolate faces in
images or video streams.
Face Alignment
detected faces are transformed into a standard format by
rotating, scaling, and translating them for uniformity.
Feature Extraction
machine learning models, such as CNNs, extract unique facial
features and create a face embedding or feature vector.
Face Matching
extracted features are compared with stored feature vectors
using distance metrics like Euclidean or cosine distance.
Systems identify the closest match or verify if the face matches
a specific representation.
Decision Making
the system determines whether to accept or reject
identification or verification based on matching results,
sometimes providing a confidence score or probability.
13
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Face Detection
algorithms like Haar cascades or SSD locate and isolate faces in
images or video streams.
Face Alignment
detected faces are transformed into a standard format by
rotating, scaling, and translating them for uniformity.
Feature Extraction
machine learning models, such as CNNs, extract unique facial
features and create a face embedding or feature vector.
Face Matching
extracted features are compared with stored feature vectors
using distance metrics like Euclidean or cosine distance.
Systems identify the closest match or verify if the face matches
a specific representation.
Decision Making
the system determines whether to accept or reject
identification or verification based on matching results,
sometimes providing a confidence score or probability.
14
Physical Biometric
Modalities
Fingerprint Recognition
Face Recognition
Iris Recognition
Retina Recognition
Hand Geometry
Vein Recognition
Ear Recognition
DNA Biometrics
Behavioral Biometric
Modalities
Voice Recognition
Signature Recognition
Keystroke Dynamics
Gait Recognition
Mouse Dynamics
Face Detection
algorithms like Haar cascades or SSD locate and isolate faces in
images or video streams.
Face Alignment
detected faces are transformed into a standard format by
rotating, scaling, and translating them for uniformity.
Feature Extraction
machine learning models, such as CNNs, extract unique facial
features and create a face embedding or feature vector.
Face Matching
extracted features are compared with stored feature vectors
using distance metrics like Euclidean or cosine distance.
Systems identify the closest match or verify if the face matches
a specific representation.
Decision Making
the system determines whether to accept or reject
identification or verification based on matching results,
sometimes providing a confidence score or probability.
15
ML attacks landscape
16
ML attacks landscape v1
AI App Security Risk
Model
Security
• Adversarial ML
• Model
Backdoor
• Model Theft
Implementation
Security
• Sensor Security
• Flaws in Framework
• Logical Flaws
Data Integrity
Security
• Data Poisoning
• Scaling Attack
• Risk over Network
https://tinyurl.com/4fh7j3ky
17
https://tinyurl.com/339uetbz
18
AI Attacks
Promt
injection
Training
attacks
AI Agents Tools Storage Models
# alter agent routing
# send commands to
undefined systems
# execute arbitrary
commands on backend
business systems
# pass through injection on
connected tool systems
# code execution on agent
system
# attack embedding
databases
# extract sensitive data
# modify embedding data
resulting in tampered model
results
# bypass model protections
# force model to exhibit bias
# extraction of other users' and/or
backend data
# force model to exhibit intolerant
behavior
# poison other users' results
# disrupt model trust/reliability
#access unpublished models
# introduce bias into
the model
# disrupt model
trust/reliability
ML attacks landscape v2
Biometric attacks landscape
19
Data
acquisition
Feature
Extraction
Face Matching Decision
Data Storage
Attack on the
sensor by biometric
presentation type
Sample
replacement
Attack
on the signal
processor
Pattern
replacement
Attack on the
comparison
algorithm
Value
replacement
Decision
replacement
Replacement of
sample (pattern)
Replacement of link
to sample
Biometrics
attack
Infrastructure attacks
Attacking devices
20
Devices overview
facial recognition
access control device
time control
device
smart doorbell
21
Device #1
1) The customer bought an expensive B2B device
which we audited in their work environment
2) Typically, multiple devices are ordered for the
project:
one — for physical hacking, the second — for
logical and testing, the third is a backup
3) The result of the physical audit. Categories of
cameras in systems and in our system. The reason
for using depth cameras
22
Overview depth camera 2 x visible
light camera
23
Assumption #1
How does it work?
1) Detecting a face in
the frame.
2) Checking Liveness
with the depth
camera.
3) Capture the face from
the visible range
camera.
4) Pre-processing.
5) DNN
6) Comparison with the
database using
threshold 2500
depth camera 2 x visible
light camera
24
Assumption #1
How does it work?
1) Detecting a face in
the frame.
2) Checking Liveness
with the depth
camera.
3) Capture the face
from the visible
range camera.
4) Pre-processing.
5) DNN
6) Comparison with the
database using
threshold 2500
25
What if there are multiple faces in the
frame?
The larger head is the one being analyzed.
Assumption #2 | Multiple faces
26
Assumption #3 | universal face?
You need to pass 2500 threshold to get access.
Hypothesis — It is possible to authenticate
without having a photo of the reference user.
Create a generated face and present it to the
system via a spoofed channel.
27
28
Assumption #3 | universal face?
Variational
Autoencoder
CelebA Dataset
Face Super-
Resolution
model
score > 2500 ?
digital physical
NO YES
Results #1
● The study unveils inadequate utilization of depth
camera data by the vendor.
● This deficiency may stem from hardware limitations,
potentially rendering the system more vulnerable to
attacks. Deep learning models do not interact with
depth maps in any way.
● Incorporating depth data in the training process
could enhance system reliability.
● However, it may also introduce complexities in the
preparation of training datasets.
29
Device #2 (ZKTeco)
1) Time tracking terminal
2) No CUDA
3) ML algorithms from 2010
30
Overview
It uses only infrared camera
31
How it works
32
Biometrical algorithms:
1) Gabor Filters https://t.co/CBFKums9TO
2) Local Binary Pattern https://t.co/OxYFkTZTP0
Gabor filter
Local binary pattern
As seen by the infrared
light camera
LED lamp inspiration
33
LED lamps emit a lot of their
energy in the form of
infrared light
LED lamp inspiration
34
printing a photo on transparent film
LED lamps emit a lot of their
energy in the form of
infrared light
LED lamp inspiration
35
LED lamps emit a lot of their
energy in the form of
infrared light
printing a photo on transparent film
shining an
incandescent light
through it
Results #2
● We discovered logical vulnerabilities in the terminal,
enabling a more detailed examination of its
functioning.
● One notable attempt involved creating a unique
single-frame screen displayed on transparent film
and illuminated with infrared light
● Unfortunately, the terminal exhibited high sensitivity
to specific changes. For instance, it identified the
same user differently when wearing or not wearing
glasses, treating them as distinct individuals.
● Nevertheless, the combination of technologies,
including Gabor filters, local binary patterns, and an
infrared camera, provides a solid defense against
potential attacks
36
Device #3 (Eufy)
Smart doorbells become the part of everyone’s life
Vendors add “AI” to the device
Now the product is more complex
Is it more secure now?
37
Overview ● The Smart Doorbell is a high-tech home security device.
It offers HD video, two-way audio, motion detection, and
local storage (c)
● It's privacy-focused with robust encryption and
integrates with other devices (c)
38
Issue #1: Man-in-the-middle attack
Device checks for firmware
updates every time it boots
There’s no SSL pinning
Firmware is “signed” with MD5
39
Issue #2: Military grade encryption
● All videos are stored on a 4GB “smart
hub”
● There’s AES-128 encryption
● Key is generate using srand() PRNG
● Seed is time()
● 30s to find the key and decrypt the
videos
40
Issue #3: Authorisation bypass
Every snapshot is
uploaded to AWS
Server generates AWS
signature for
uploading/downloading
41
Issue #3: Authorisation bypass
Every snapshot is uploaded to AWS
Server generates AWS signature for
uploading/downloading
Path traversal in link signature generation
Any snapshot of any eufy user is available
42
Issue #4: Unlocked USB-OTG
Direct physical access to shell
Access to firmware binaries
model.bin.tar
43
Overview
● The Smart Doorbell is a high-tech home security device.
It offers HD video, two-way audio, motion detection, and
local storage (c)
● It's privacy-focused with robust encryption and
integrates with other devices (c)
● You can choose between battery or wired installation,
and it's weather-resistant. Control it via a user-friendly
app for remote monitoring and alerts (c)
44
Overview
● The Smart Doorbell is a high-tech home security device.
It offers HD video, two-way audio, motion detection, and
local storage (c)
● It's privacy-focused with robust encryption and
integrates with other devices (c)
● You can choose between battery or wired installation,
and it's weather-resistant. Control it via a user-friendly
app for remote monitoring and alerts (c)
45
Is it still vulnerable?
46
https://github.com/kripthor/talks_and_slides/blob
/main/IoT-Landscape.pdf
47
More evidence that Eufy can’t be
hacked
48
Lessons learned
Newer, better, more secure - False
More advanced ML - more resilient algorithms - False
Cheaper devices - less security - False
49
Checklist
50
Hardware/Software
- Enumerate interfaces
- ethernet
- USB, serial and debugging ports
- mics and cameras
- Investigate available cameras
- infra-red, depth camera, etc
- Firmware
- Download the FW from public or using MiTM
- Open a device and extract the FW from a chip
- Get information about the vendor
- Can the models and algorithms be extracted
- Where and how images/videos are stored and processed (cloud or on-prem)
- Assess the infrastructure and public libs
Data privacy & Model robustness (Grey Box)
- Errors in the recognition pipeline
- Adversarial attacks
- deepfakes
- universal faces
- similar faces
- Liveness checks
Data integrity & Model confidentiality tests (Black Box)
- Interfering with sensors
- With light
- By the channel interference
- Spoofing
- Determine crucial elements on a face by overlapping parts
- Can we use a digital face instead, e.g., a large LCD
- DDoS by presenting a large number of faces
- Applying patches and masks
- Data stealing
- Targeted and untargeted attacks
Kudos
51
Alexander Migutsky
Denis Goryushev
Egor Zaitsev
Dmitry Sklyarov
Pedro Umbelino
Cyber R&D Lab (RIP)
1 de 51

Recomendados

Deepfakes: Trick or Treat? por
Deepfakes: Trick or Treat?Deepfakes: Trick or Treat?
Deepfakes: Trick or Treat?Ian McCarthy
15.4K visualizações28 slides
Deepfakes por
DeepfakesDeepfakes
DeepfakesCA.Kolluru Narayanarao
404 visualizações8 slides
"Creating, Weaponizing,and Detecting Deep Fakes," a Presentation from U.C. Be... por
"Creating, Weaponizing,and Detecting Deep Fakes," a Presentation from U.C. Be..."Creating, Weaponizing,and Detecting Deep Fakes," a Presentation from U.C. Be...
"Creating, Weaponizing,and Detecting Deep Fakes," a Presentation from U.C. Be...Edge AI and Vision Alliance
1.1K visualizações37 slides
DeepFake Detection: Challenges, Progress and Hands-on Demonstration of Techno... por
DeepFake Detection: Challenges, Progress and Hands-on Demonstration of Techno...DeepFake Detection: Challenges, Progress and Hands-on Demonstration of Techno...
DeepFake Detection: Challenges, Progress and Hands-on Demonstration of Techno...Symeon Papadopoulos
889 visualizações29 slides
Deep fake por
Deep fakeDeep fake
Deep fakeelisetty pavan
1.3K visualizações19 slides
DeepFake: Trick or Treat por
DeepFake: Trick or TreatDeepFake: Trick or Treat
DeepFake: Trick or TreatSamama Khan
788 visualizações30 slides

Mais conteúdo relacionado

Mais procurados

Deepfakes - How they work and what it means for the future por
Deepfakes - How they work and what it means for the futureDeepfakes - How they work and what it means for the future
Deepfakes - How they work and what it means for the futureJarrod Overson
4K visualizações64 slides
Deepfake.pptx por
Deepfake.pptxDeepfake.pptx
Deepfake.pptxNandeeshNandhu2
1.4K visualizações12 slides
Face recognition technology por
Face recognition technologyFace recognition technology
Face recognition technologyShubhamLamichane
1.8K visualizações31 slides
DeepFake_Seminar.pptx por
DeepFake_Seminar.pptxDeepFake_Seminar.pptx
DeepFake_Seminar.pptxsandeshsb
287 visualizações20 slides
Deep fakes and beyond por
Deep fakes and beyondDeep fakes and beyond
Deep fakes and beyondArpitha Gurumurthy
214 visualizações29 slides
chatgpt-privacy and security.pptx por
chatgpt-privacy and security.pptxchatgpt-privacy and security.pptx
chatgpt-privacy and security.pptxDeepak Kumar
251 visualizações23 slides

Mais procurados(20)

Deepfakes - How they work and what it means for the future por Jarrod Overson
Deepfakes - How they work and what it means for the futureDeepfakes - How they work and what it means for the future
Deepfakes - How they work and what it means for the future
Jarrod Overson4K visualizações
Deepfake.pptx por NandeeshNandhu2
Deepfake.pptxDeepfake.pptx
Deepfake.pptx
NandeeshNandhu21.4K visualizações
Face recognition technology por ShubhamLamichane
Face recognition technologyFace recognition technology
Face recognition technology
ShubhamLamichane1.8K visualizações
DeepFake_Seminar.pptx por sandeshsb
DeepFake_Seminar.pptxDeepFake_Seminar.pptx
DeepFake_Seminar.pptx
sandeshsb287 visualizações
Deep fakes and beyond por Arpitha Gurumurthy
Deep fakes and beyondDeep fakes and beyond
Deep fakes and beyond
Arpitha Gurumurthy214 visualizações
chatgpt-privacy and security.pptx por Deepak Kumar
chatgpt-privacy and security.pptxchatgpt-privacy and security.pptx
chatgpt-privacy and security.pptx
Deepak Kumar251 visualizações
Exploring VR por FITC
 Exploring VR Exploring VR
Exploring VR
FITC24.7K visualizações
Deception technology for advanced detection por Jisc
Deception technology for advanced detectionDeception technology for advanced detection
Deception technology for advanced detection
Jisc2.2K visualizações
Securing the Cloud por GGV Capital
Securing the CloudSecuring the Cloud
Securing the Cloud
GGV Capital40.8K visualizações
Cyber security threats for 2017 por Ramiro Cid
Cyber security threats for 2017Cyber security threats for 2017
Cyber security threats for 2017
Ramiro Cid26K visualizações
Face recognition system por Transistor Amplifier
Face recognition systemFace recognition system
Face recognition system
Transistor Amplifier1.7K visualizações
Face Recognition por arvind carpenter
Face RecognitionFace Recognition
Face Recognition
arvind carpenter1.3K visualizações
Facial powerpoint por 12202843
Facial powerpointFacial powerpoint
Facial powerpoint
122028432.8K visualizações
Deep Fakes Artificial Intelligence.pptx por NilayDeshmukh3
Deep Fakes Artificial Intelligence.pptxDeep Fakes Artificial Intelligence.pptx
Deep Fakes Artificial Intelligence.pptx
NilayDeshmukh3206 visualizações
Face recognition application por awadhesh kumar
Face recognition applicationFace recognition application
Face recognition application
awadhesh kumar1.4K visualizações
FACE RECOGNITION TECHNOLOGY por JASHU JASWANTH
FACE RECOGNITION TECHNOLOGYFACE RECOGNITION TECHNOLOGY
FACE RECOGNITION TECHNOLOGY
JASHU JASWANTH39.1K visualizações
Deepfake detection por Weverify
Deepfake detectionDeepfake detection
Deepfake detection
Weverify5.8K visualizações
Face detection and recognition por Pankaj Thakur
Face detection and recognitionFace detection and recognition
Face detection and recognition
Pankaj Thakur13.6K visualizações
Deepfakes: An Emerging Internet Threat and their Detection por Symeon Papadopoulos
Deepfakes: An Emerging Internet Threat and their DetectionDeepfakes: An Emerging Internet Threat and their Detection
Deepfakes: An Emerging Internet Threat and their Detection
Symeon Papadopoulos1.5K visualizações
Face recognition por sandeepsharma1193
Face recognitionFace recognition
Face recognition
sandeepsharma11934.4K visualizações

Similar a Unlocking Any Door In The 21st Century. Immersion In Biometric Security.

biometric technology por
biometric technologybiometric technology
biometric technologyAnmol Bagga
18.6K visualizações33 slides
Pattern recognition 3d face recognition por
Pattern recognition 3d face recognitionPattern recognition 3d face recognition
Pattern recognition 3d face recognitionMazin Alwaaly
932 visualizações30 slides
Facial recognition por
Facial recognitionFacial recognition
Facial recognitionDhimankomal
1.3K visualizações20 slides
Attendance system based on face recognition using python by Raihan Sikdar por
Attendance system based on face recognition using python by Raihan SikdarAttendance system based on face recognition using python by Raihan Sikdar
Attendance system based on face recognition using python by Raihan Sikdarraihansikdar
6.4K visualizações32 slides
Facial Recognition System por
Facial Recognition SystemFacial Recognition System
Facial Recognition SystemArun ACE
6.6K visualizações30 slides
Facial recognition technology by vaibhav por
Facial recognition technology by vaibhavFacial recognition technology by vaibhav
Facial recognition technology by vaibhavVaibhav P
7.1K visualizações34 slides

Similar a Unlocking Any Door In The 21st Century. Immersion In Biometric Security.(20)

biometric technology por Anmol Bagga
biometric technologybiometric technology
biometric technology
Anmol Bagga18.6K visualizações
Pattern recognition 3d face recognition por Mazin Alwaaly
Pattern recognition 3d face recognitionPattern recognition 3d face recognition
Pattern recognition 3d face recognition
Mazin Alwaaly932 visualizações
Facial recognition por Dhimankomal
Facial recognitionFacial recognition
Facial recognition
Dhimankomal1.3K visualizações
Attendance system based on face recognition using python by Raihan Sikdar por raihansikdar
Attendance system based on face recognition using python by Raihan SikdarAttendance system based on face recognition using python by Raihan Sikdar
Attendance system based on face recognition using python by Raihan Sikdar
raihansikdar6.4K visualizações
Facial Recognition System por Arun ACE
Facial Recognition SystemFacial Recognition System
Facial Recognition System
Arun ACE6.6K visualizações
Facial recognition technology by vaibhav por Vaibhav P
Facial recognition technology by vaibhavFacial recognition technology by vaibhav
Facial recognition technology by vaibhav
Vaibhav P7.1K visualizações
Chapters 3 4 por sakshi_20
Chapters 3 4Chapters 3 4
Chapters 3 4
sakshi_201K visualizações
International Journal of Engineering and Science Invention (IJESI) por inventionjournals
International Journal of Engineering and Science Invention (IJESI)International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)
inventionjournals327 visualizações
Face Recognition Technology por Shashidhar Reddy
Face Recognition TechnologyFace Recognition Technology
Face Recognition Technology
Shashidhar Reddy783 visualizações
Biometric technology por Sudip Sadhukhan
Biometric technologyBiometric technology
Biometric technology
Sudip Sadhukhan1.1K visualizações
Scale Invariant Feature Transform Based Face Recognition from a Single Sample... por ijceronline
Scale Invariant Feature Transform Based Face Recognition from a Single Sample...Scale Invariant Feature Transform Based Face Recognition from a Single Sample...
Scale Invariant Feature Transform Based Face Recognition from a Single Sample...
ijceronline377 visualizações
Bio shiv por Shivank Garg
Bio shivBio shiv
Bio shiv
Shivank Garg183 visualizações
G041041047 por ijceronline
G041041047G041041047
G041041047
ijceronline266 visualizações
Biometrics por Shivani Gautam
BiometricsBiometrics
Biometrics
Shivani Gautam1.3K visualizações
76 s201920 por IJRAT
76 s20192076 s201920
76 s201920
IJRAT20 visualizações
Face recognition technology por Pushkar Dutt
Face recognition technologyFace recognition technology
Face recognition technology
Pushkar Dutt724 visualizações
Ijetcas14 598 por Iasir Journals
Ijetcas14 598Ijetcas14 598
Ijetcas14 598
Iasir Journals299 visualizações
Biometrics Research/Thesis Paper por Sumaiya Ismail
Biometrics Research/Thesis PaperBiometrics Research/Thesis Paper
Biometrics Research/Thesis Paper
Sumaiya Ismail116 visualizações

Mais de Payment Village

Offensive Payment Security por
Offensive Payment SecurityOffensive Payment Security
Offensive Payment SecurityPayment Village
2.5K visualizações94 slides
The white whales of fraud and where to find them por
The white whales of fraud and where to find themThe white whales of fraud and where to find them
The white whales of fraud and where to find themPayment Village
41 visualizações21 slides
Breaking banks or saving them por
Breaking banks or saving themBreaking banks or saving them
Breaking banks or saving themPayment Village
142 visualizações31 slides
Hack in Cash out OWASP London por
Hack in Cash out OWASP LondonHack in Cash out OWASP London
Hack in Cash out OWASP LondonPayment Village
330 visualizações38 slides
Ради денег. Безопасность платежных терминалов por
Ради денег. Безопасность платежных терминаловРади денег. Безопасность платежных терминалов
Ради денег. Безопасность платежных терминаловPayment Village
155 visualizações20 slides
Launch Impossible Current State of Application Control Bypasses on ATMs. por
Launch Impossible Current State of Application Control Bypasses on ATMs.Launch Impossible Current State of Application Control Bypasses on ATMs.
Launch Impossible Current State of Application Control Bypasses on ATMs.Payment Village
372 visualizações60 slides

Mais de Payment Village(8)

Offensive Payment Security por Payment Village
Offensive Payment SecurityOffensive Payment Security
Offensive Payment Security
Payment Village2.5K visualizações
The white whales of fraud and where to find them por Payment Village
The white whales of fraud and where to find themThe white whales of fraud and where to find them
The white whales of fraud and where to find them
Payment Village41 visualizações
Breaking banks or saving them por Payment Village
Breaking banks or saving themBreaking banks or saving them
Breaking banks or saving them
Payment Village142 visualizações
Hack in Cash out OWASP London por Payment Village
Hack in Cash out OWASP LondonHack in Cash out OWASP London
Hack in Cash out OWASP London
Payment Village330 visualizações
Ради денег. Безопасность платежных терминалов por Payment Village
Ради денег. Безопасность платежных терминаловРади денег. Безопасность платежных терминалов
Ради денег. Безопасность платежных терминалов
Payment Village155 visualizações
Launch Impossible Current State of Application Control Bypasses on ATMs. por Payment Village
Launch Impossible Current State of Application Control Bypasses on ATMs.Launch Impossible Current State of Application Control Bypasses on ATMs.
Launch Impossible Current State of Application Control Bypasses on ATMs.
Payment Village372 visualizações
How to build Big Brother por Payment Village
How to build Big BrotherHow to build Big Brother
How to build Big Brother
Payment Village610 visualizações
СМС – «золотой» стандарт двухфакторной аутентификации. Актуальные проблемы por Payment Village
СМС – «золотой» стандарт двухфакторной аутентификации. Актуальные проблемыСМС – «золотой» стандарт двухфакторной аутентификации. Актуальные проблемы
СМС – «золотой» стандарт двухфакторной аутентификации. Актуальные проблемы
Payment Village527 visualizações

Unlocking Any Door In The 21st Century. Immersion In Biometric Security.

  • 1. Unlocking any door in the 21st century Immersion in biometric security 1 Timur Yunusov & Alexandra Murzina
  • 2. Who we are ● ex A-Team Cyber R&D Lab ● Head of research ● Senior ML security expert 2
  • 3. Outline ● Current state of AI/ML in biometrics ● ML attacks landscape ● Attacking devices ○ Device 1 - undisclosed ○ Device 2 - ZKTeco ○ Device 3 - Eufy ● Conclusions ● Security Checklist 3
  • 4. United States: State-specific biometric laws, e.g., BIPA in Illinois and CCPA in California. FBI uses biometrics for law enforcement and border control. China: Extensive government use of biometrics for surveillance and security. Requirement to store critical data, including biometrics, within the country. India: Aadhaar Act regulates biometric data collected under Aadhaar program. Proposed Data Privacy Bill aims for comprehensive data protection. European Union (EU): GDPR regulates biometric data with explicit consent and stringent protection. United Arab Emirates (UAE): DIFC's data protection law covers biometric data. Government uses biometrics extensively for security and services. Japan: APPI regulates personal data, including biometrics, with consent and protection. Legislation United Kingdom: Data Protection Act regulates personal data processing, including biometrics. Independent oversight of law enforcement biometric use by Biometrics Commissioner. South Korea: PIPA considers biometric data "sensitive," requiring consent and protection. Regulations allow biometric authentication in financial transactions. Brazil: LGPD regulates personal data processing, including biometrics, with consent and protection. Requires security measures and impact assessments. South Africa: POPIA regulates personal data processing, including biometrics, with consent and protection. Russia: Personal Data Law mandates consent for biometric processing. Federal Law regulates fingerprinting. Unified Biometric System enables bank identification. Government uses biometrics for security and law enforcement. 4
  • 6. Current state of AI/ML in biometrics 6
  • 7. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics 7
  • 8. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Early Methods, Eigenfaces initially, manual analysis of facial features in photos measured distances and angles between landmarks like eyes and nose. Automated face recognition began in the late 1980s with Eigenfaces, using PCA to extract features from grayscale images, representing faces as weighted "eigenfaces." Local Feature Methods techniques like LBP and Gabor wavelets focused on specific face regions, capturing texture and local changes. 2D and 3D Face Models 2D and 3D face models accounted for pose and expression variations, with 3D models providing depth information. 8
  • 9. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Early Methods, Eigenfaces initially, manual analysis of facial features in photos measured distances and angles between landmarks like eyes and nose. Automated face recognition began in the late 1980s with Eigenfaces, using PCA to extract features from grayscale images, representing faces as weighted "eigenfaces." Local Feature Methods techniques like LBP and Gabor wavelets focused on specific face regions, capturing texture and local changes. 2D and 3D Face Models 2D and 3D face models accounted for pose and expression variations, with 3D models providing depth information. Machine Learning and Deep Learning machine learning and deep learning techniques, like SVMs and CNNs, automatically learned and extracted facial features from large datasets, enhancing recognition accuracy and robustness. 9
  • 10. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Early Methods, Eigenfaces initially, manual analysis of facial features in photos measured distances and angles between landmarks like eyes and nose. Automated face recognition began in the late 1980s with Eigenfaces, using PCA to extract features from grayscale images, representing faces as weighted "eigenfaces." Local Feature Methods techniques like LBP and Gabor wavelets focused on specific face regions, capturing texture and local changes. 2D and 3D Face Models 2D and 3D face models accounted for pose and expression variations, with 3D models providing depth information. Machine Learning and Deep Learning machine learning and deep learning techniques, like SVMs and CNNs, automatically learned and extracted facial features from large datasets, enhancing recognition accuracy and robustness. Depth Sensing and Infrared Cameras Modern systems use depth sensing and infrared cameras to capture facial information in challenging lighting or obscured faces, enabling accurate recognition and spoof detection. Multi-modal and Fusion Methods Combining multiple biometric modalities, such as face and voice or fusing 2D and 3D data, has enhanced recognition performance. Emotion Recognition and Liveness Detection Recent advancements include emotion recognition from facial expressions and liveness detection to verify the subject's presence. 10
  • 11. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Face Detection algorithms like Haar cascades or SSD locate and isolate faces in images or video streams. Face Alignment detected faces are transformed into a standard format by rotating, scaling, and translating them for uniformity. Feature Extraction machine learning models, such as CNNs, extract unique facial features and create a face embedding or feature vector. Face Matching extracted features are compared with stored feature vectors using distance metrics like Euclidean or cosine distance. Systems identify the closest match or verify if the face matches a specific representation. Decision Making the system determines whether to accept or reject identification or verification based on matching results, sometimes providing a confidence score or probability. 11
  • 12. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Face Detection algorithms like Haar cascades or SSD locate and isolate faces in images or video streams. Face Alignment detected faces are transformed into a standard format by rotating, scaling, and translating them for uniformity. Feature Extraction machine learning models, such as CNNs, extract unique facial features and create a face embedding or feature vector. Face Matching extracted features are compared with stored feature vectors using distance metrics like Euclidean or cosine distance. Systems identify the closest match or verify if the face matches a specific representation. Decision Making the system determines whether to accept or reject identification or verification based on matching results, sometimes providing a confidence score or probability. 12
  • 13. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Face Detection algorithms like Haar cascades or SSD locate and isolate faces in images or video streams. Face Alignment detected faces are transformed into a standard format by rotating, scaling, and translating them for uniformity. Feature Extraction machine learning models, such as CNNs, extract unique facial features and create a face embedding or feature vector. Face Matching extracted features are compared with stored feature vectors using distance metrics like Euclidean or cosine distance. Systems identify the closest match or verify if the face matches a specific representation. Decision Making the system determines whether to accept or reject identification or verification based on matching results, sometimes providing a confidence score or probability. 13
  • 14. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Face Detection algorithms like Haar cascades or SSD locate and isolate faces in images or video streams. Face Alignment detected faces are transformed into a standard format by rotating, scaling, and translating them for uniformity. Feature Extraction machine learning models, such as CNNs, extract unique facial features and create a face embedding or feature vector. Face Matching extracted features are compared with stored feature vectors using distance metrics like Euclidean or cosine distance. Systems identify the closest match or verify if the face matches a specific representation. Decision Making the system determines whether to accept or reject identification or verification based on matching results, sometimes providing a confidence score or probability. 14
  • 15. Physical Biometric Modalities Fingerprint Recognition Face Recognition Iris Recognition Retina Recognition Hand Geometry Vein Recognition Ear Recognition DNA Biometrics Behavioral Biometric Modalities Voice Recognition Signature Recognition Keystroke Dynamics Gait Recognition Mouse Dynamics Face Detection algorithms like Haar cascades or SSD locate and isolate faces in images or video streams. Face Alignment detected faces are transformed into a standard format by rotating, scaling, and translating them for uniformity. Feature Extraction machine learning models, such as CNNs, extract unique facial features and create a face embedding or feature vector. Face Matching extracted features are compared with stored feature vectors using distance metrics like Euclidean or cosine distance. Systems identify the closest match or verify if the face matches a specific representation. Decision Making the system determines whether to accept or reject identification or verification based on matching results, sometimes providing a confidence score or probability. 15
  • 17. ML attacks landscape v1 AI App Security Risk Model Security • Adversarial ML • Model Backdoor • Model Theft Implementation Security • Sensor Security • Flaws in Framework • Logical Flaws Data Integrity Security • Data Poisoning • Scaling Attack • Risk over Network https://tinyurl.com/4fh7j3ky 17
  • 18. https://tinyurl.com/339uetbz 18 AI Attacks Promt injection Training attacks AI Agents Tools Storage Models # alter agent routing # send commands to undefined systems # execute arbitrary commands on backend business systems # pass through injection on connected tool systems # code execution on agent system # attack embedding databases # extract sensitive data # modify embedding data resulting in tampered model results # bypass model protections # force model to exhibit bias # extraction of other users' and/or backend data # force model to exhibit intolerant behavior # poison other users' results # disrupt model trust/reliability #access unpublished models # introduce bias into the model # disrupt model trust/reliability ML attacks landscape v2
  • 19. Biometric attacks landscape 19 Data acquisition Feature Extraction Face Matching Decision Data Storage Attack on the sensor by biometric presentation type Sample replacement Attack on the signal processor Pattern replacement Attack on the comparison algorithm Value replacement Decision replacement Replacement of sample (pattern) Replacement of link to sample Biometrics attack Infrastructure attacks
  • 21. Devices overview facial recognition access control device time control device smart doorbell 21
  • 22. Device #1 1) The customer bought an expensive B2B device which we audited in their work environment 2) Typically, multiple devices are ordered for the project: one — for physical hacking, the second — for logical and testing, the third is a backup 3) The result of the physical audit. Categories of cameras in systems and in our system. The reason for using depth cameras 22
  • 23. Overview depth camera 2 x visible light camera 23
  • 24. Assumption #1 How does it work? 1) Detecting a face in the frame. 2) Checking Liveness with the depth camera. 3) Capture the face from the visible range camera. 4) Pre-processing. 5) DNN 6) Comparison with the database using threshold 2500 depth camera 2 x visible light camera 24
  • 25. Assumption #1 How does it work? 1) Detecting a face in the frame. 2) Checking Liveness with the depth camera. 3) Capture the face from the visible range camera. 4) Pre-processing. 5) DNN 6) Comparison with the database using threshold 2500 25
  • 26. What if there are multiple faces in the frame? The larger head is the one being analyzed. Assumption #2 | Multiple faces 26
  • 27. Assumption #3 | universal face? You need to pass 2500 threshold to get access. Hypothesis — It is possible to authenticate without having a photo of the reference user. Create a generated face and present it to the system via a spoofed channel. 27
  • 28. 28 Assumption #3 | universal face? Variational Autoencoder CelebA Dataset Face Super- Resolution model score > 2500 ? digital physical NO YES
  • 29. Results #1 ● The study unveils inadequate utilization of depth camera data by the vendor. ● This deficiency may stem from hardware limitations, potentially rendering the system more vulnerable to attacks. Deep learning models do not interact with depth maps in any way. ● Incorporating depth data in the training process could enhance system reliability. ● However, it may also introduce complexities in the preparation of training datasets. 29
  • 30. Device #2 (ZKTeco) 1) Time tracking terminal 2) No CUDA 3) ML algorithms from 2010 30
  • 31. Overview It uses only infrared camera 31
  • 32. How it works 32 Biometrical algorithms: 1) Gabor Filters https://t.co/CBFKums9TO 2) Local Binary Pattern https://t.co/OxYFkTZTP0 Gabor filter Local binary pattern As seen by the infrared light camera
  • 33. LED lamp inspiration 33 LED lamps emit a lot of their energy in the form of infrared light
  • 34. LED lamp inspiration 34 printing a photo on transparent film LED lamps emit a lot of their energy in the form of infrared light
  • 35. LED lamp inspiration 35 LED lamps emit a lot of their energy in the form of infrared light printing a photo on transparent film shining an incandescent light through it
  • 36. Results #2 ● We discovered logical vulnerabilities in the terminal, enabling a more detailed examination of its functioning. ● One notable attempt involved creating a unique single-frame screen displayed on transparent film and illuminated with infrared light ● Unfortunately, the terminal exhibited high sensitivity to specific changes. For instance, it identified the same user differently when wearing or not wearing glasses, treating them as distinct individuals. ● Nevertheless, the combination of technologies, including Gabor filters, local binary patterns, and an infrared camera, provides a solid defense against potential attacks 36
  • 37. Device #3 (Eufy) Smart doorbells become the part of everyone’s life Vendors add “AI” to the device Now the product is more complex Is it more secure now? 37
  • 38. Overview ● The Smart Doorbell is a high-tech home security device. It offers HD video, two-way audio, motion detection, and local storage (c) ● It's privacy-focused with robust encryption and integrates with other devices (c) 38
  • 39. Issue #1: Man-in-the-middle attack Device checks for firmware updates every time it boots There’s no SSL pinning Firmware is “signed” with MD5 39
  • 40. Issue #2: Military grade encryption ● All videos are stored on a 4GB “smart hub” ● There’s AES-128 encryption ● Key is generate using srand() PRNG ● Seed is time() ● 30s to find the key and decrypt the videos 40
  • 41. Issue #3: Authorisation bypass Every snapshot is uploaded to AWS Server generates AWS signature for uploading/downloading 41
  • 42. Issue #3: Authorisation bypass Every snapshot is uploaded to AWS Server generates AWS signature for uploading/downloading Path traversal in link signature generation Any snapshot of any eufy user is available 42
  • 43. Issue #4: Unlocked USB-OTG Direct physical access to shell Access to firmware binaries model.bin.tar 43
  • 44. Overview ● The Smart Doorbell is a high-tech home security device. It offers HD video, two-way audio, motion detection, and local storage (c) ● It's privacy-focused with robust encryption and integrates with other devices (c) ● You can choose between battery or wired installation, and it's weather-resistant. Control it via a user-friendly app for remote monitoring and alerts (c) 44
  • 45. Overview ● The Smart Doorbell is a high-tech home security device. It offers HD video, two-way audio, motion detection, and local storage (c) ● It's privacy-focused with robust encryption and integrates with other devices (c) ● You can choose between battery or wired installation, and it's weather-resistant. Control it via a user-friendly app for remote monitoring and alerts (c) 45
  • 46. Is it still vulnerable? 46
  • 48. More evidence that Eufy can’t be hacked 48
  • 49. Lessons learned Newer, better, more secure - False More advanced ML - more resilient algorithms - False Cheaper devices - less security - False 49
  • 50. Checklist 50 Hardware/Software - Enumerate interfaces - ethernet - USB, serial and debugging ports - mics and cameras - Investigate available cameras - infra-red, depth camera, etc - Firmware - Download the FW from public or using MiTM - Open a device and extract the FW from a chip - Get information about the vendor - Can the models and algorithms be extracted - Where and how images/videos are stored and processed (cloud or on-prem) - Assess the infrastructure and public libs Data privacy & Model robustness (Grey Box) - Errors in the recognition pipeline - Adversarial attacks - deepfakes - universal faces - similar faces - Liveness checks Data integrity & Model confidentiality tests (Black Box) - Interfering with sensors - With light - By the channel interference - Spoofing - Determine crucial elements on a face by overlapping parts - Can we use a digital face instead, e.g., a large LCD - DDoS by presenting a large number of faces - Applying patches and masks - Data stealing - Targeted and untargeted attacks
  • 51. Kudos 51 Alexander Migutsky Denis Goryushev Egor Zaitsev Dmitry Sklyarov Pedro Umbelino Cyber R&D Lab (RIP)