Facial Recognition System

Meet

DSBio

Biometric verification for a more secure space. With fast and accurate face recognition, liveness detection, face matching, and identification, ensure that only real people can access your services.

What is DSBio?

We provide DSBio Face Recognition System (FRS) as an API with RESTful services. This allows for very quick and easy integration with any external system.

Implemented in Python and already available as Web Services, DSBio is also available as a C++ binary package to be integrated on several architectures: Android, iOS, PCs, etc.

DSBio is the solution we found for our clients who were looking for a simple face recognition system that would get the job done without too much hassle. We like to think of our system as a complex project turned into a simple solution. DSBio can provide everything you are looking for in a biometric verification system, minding your requirements and resources.

The possibilities with DSBio

The functionalities supported by DSBio are:

Enroll a person (or identity)

Given an image, the system will register the person in the biometric database and return the assigned ID. Additionally, the functionality allows receiving the ID to be generated from the external system (for example, the ID could be the identity document number).

Note: Before enrollment, a search is performed throughout the biometric database to determine if the same face was previously registered with another identity. Enrollment is not allowed if a similar record is found. This mechanism keeps the biometric database "clean" of duplicates, regardless of the logic implemented by the external system.

Image identification

Given an image, the system will search the entire database and return the ID(s) of the most similar person(s).

Add an image to a person

Given an ID and a new image, the system will add this new record to the data associated with that person.

Delete a person

Given an ID, the system will delete the person associated with that ID, along with all associated biometric records.

Verify a person's identity

Given an image and an ID, the system will return a score indicating how likely it is that the image belongs to the person identified by the ID. This allows the decision to be made in the external system. Optionally, a system-level threshold can be defined so that the verification result is True or False.

Real-Time Face Detection

We can provide our DS Face Detector, which enables face detection in real-time.

How does DSBio work?

Any FRS has three basic operations, and most of the analysis and comparisons between different systems can be based on them. These operations are:

Enrolment

Given an image face, compute the corresponding "feature vector" and store it within a database, with its corresponding ID.

Verification

Given an image face and an ID, compare the given image with the image(s) stored in the database for the given ID and return a score value that represents how similar the input image is to the ones stored in the database for that ID.

Identification

Given an image, find the more similar ID.

Face Detection

In addition to these three functions, there is an additional one: the ability to detect a face in an image. We will include this fourth operation and call it "Face Detection." Note that, in general, most FRS benchmarks are based on and described in a Verification scenario. Usual metrics, such as False Acceptance Rate, False Rejection Rate, and True Acceptance Rate, are all related to verification.

Performance

The purpose of a biometric system is always to have low FPR values. Our FRS not only performs better in terms of general Accuracy, but it's clearly superior in relation to the TPR for an FPR of 0.1%.

Why choose DSBio?

The most important aspect of any facial recognition system is the level of accuracy that can be achieved. This level of accuracy depends on the context, as 2D images of people's faces are highly influenced by external factors such as lighting, pose, expression, the passage of time, and occlusions.

DSBio Face Recognition System is state-of-the-art in terms of accuracy:

Why choose Digital Sense for your Biometrics Solutions?

Because of our expertise and accuracy of our solutions. We have developed several projects related to biometrics, particularly in facial and fingerprint recognition systems. Our experts in the field have over 10 years of national and international experience in public and private projects related to biometrics. 

Our clients choose us because these key factors:

Client Dedication

Our commitment to our clients led us to develop a solution just for their needs. DSBio is rooted in our determination to provide the best possible service.

Unmatched Expertise

Our team comprises highly experienced engineers and researchers with a deep understanding of biometric technologies. We have published several scientific publications and even a best practices manual published by the Inter-American Development Bank (IDB). 

Leading and Integrative Algorithms

Our algorithms make all the difference. They run fast and accurately and can easily integrate into any platform and hardware.

Unwavering Commitment to Security

Data privacy and security are paramount to us. We implement industry-leading security protocols to safeguard your biometric data.

DSBio at Uruguay’s Airport

Our solution goes beyond frontiers. It allows all visitors and travelers in and out of our country. Check it out.

Onboarding with Digital Sense

To integrate our facial recognition system with you, we take the approach of the Inter-American Development Bank.

1

Context Assessment

In this initial phase, it's crucial to establish a clear understanding of the specific goals and objectives for integrating DSBio into your IDS. This involves evaluating the capabilities of your existing systems and identifying areas where DSBio can add value. Additionally, assessing the availability and quality of facial image data is essential for training and testing the system effectively.

2

DSBio Fitting

Once the context is established, the next step is integrating DSBio into your solution. This involves connecting the systems through APIs, mapping data between the two platforms, and configuring DSBio settings to match your specific requirements. Customization options may include adjusting performance thresholds, accuracy levels, and privacy settings to optimize the system for your specific use case.

3

Ethical & Legal Understanding

The final step involves carefully considering facial recognition technology's ethical and legal implications. Potential biases, discrimination risks, and privacy concerns associated with such systems must be evaluated. Compliance with relevant laws and regulations, such as data privacy laws and anti-discrimination legislation, is also crucial. Conducting a privacy impact assessment can help identify and mitigate potential risks to individuals' privacy.

Answer Your Questions Today

FAQs

What you need to know about Biometric Consulting Services.

How much does working with a biometric consulting agency cost?

The answer to this question is not black or white. Unfortunately, we can’t tell you how much the consultancy or advisory can cost as we have to estimate the complexity of your project. You can always contact us and tell us about your needs. In 30 minutes, you can have a clearer view of your next steps with biometrics.

How long are biometric projects?

Biometrics can be a complex project to tackle. Nonetheless, we can estimate timeframes, costs, resources needed, and everything you need to complete it. Digital Sense will provide a team of experts who can determine a clear roadmap for you within your estimated times.

Are biometrics better than passwords?

Each system works differently and presents benefits and disadvantages. Nonetheless, we do believe that the fact that you can forget a password and create a new one makes biometrics safer for authenticating the user, providing more safety.


Can biometrics be easily hacked?

Biometric systems are not immune to hackers, yet it way more difficult to do so. Every biometric system works with a set of false positives or negatives and each system has different rates, which makes them more or less safe for users. You can check our article “Can we really fool face recognition?” for more information.

Start writing yours today

View more Success Stories
No items found.
previous
next
Latest articles

Dive Deeper into Digital Sense Knowledge

View all
No items found.
View all