FaceMe offers two main product categories:
SDK & API: FaceMe SDK and FaceMe Platform enable developers and system integrators to integrate facial recognition and related AI capabilities into existing products, systems, and services.
Turnkey Solutions: FaceMe Security, People Tracker, and FaceMe eKYC are ready-to-deploy solutions for access control and security surveillance, cross-camera person search, and digital identity verification (eKYC).
For a standalone edge device or AIoT product that runs facial recognition directly on the device, choose FaceMe SDK. It supports Windows, Linux, Android, and iOS for flexible integration across different device platforms.
For server-based or multi-device applications that require centralized management of devices, recognition records, and identity data, choose FaceMe Platform. It supports both on-premise and cloud deployments, with containerized deployment available for easier scaling and maintenance.
If you don’t need to develop your own application, FaceMe Security provides a ready-to-deploy software solution for access control and security surveillance.
FaceMe Security can integrate with existing IP cameras and VMS platforms, work with Android-based facial recognition terminals, and provide centralized management of cameras and face databases.
CyberLink provides software solutions. If you need hardware recommendations or hardware-software integration, contact our sales team, and we can connect you with a suitable system integration partner based on your requirements.
FaceMe focuses on facial recognition software solutions, while hardware selection, on-site installation, and system integration are handled by our system integration partners based on each project's requirements.
If needed, the FaceMe team can recommend suitable hardware and system integration partners to help ensure a smooth deployment of the complete solution.
Yes. FaceMe products are built on the same core facial recognition engine and can be combined across different deployment scenarios and system architectures. Face templates can also be shared across FaceMe products, reducing the need for duplicate enrollment and simplifying system integration.
For example, FaceMe SDK can be integrated into an edge device for on-device facial recognition, while FaceMe Security or FaceMe Platform can be used for centralized management, matching, and alerts.
The recommended product combination and system architecture can be determined with the FaceMe sales and technical teams during the POC or solution design stage.
FaceMe supports on-device/edge, on-premises server, and cloud deployments, depending on the application and use case. The right architecture can be selected based on requirements such as latency, bandwidth, data privacy, scalability, and enterprise IT policies.
FaceMe also supports hybrid architectures, where certain facial recognition processing can be performed at the edge while required data or results are sent to a central server for further management and applications.
Start by reviewing independent benchmark results such as NIST FRTE to objectively compare facial recognition accuracy. FaceMe achieved a 99.83% TAR at an FMR of 1E-6 in NIST FRTE 1:1 face verification, and 99.61% accuracy in a 12-million-face database search in NIST FRTE 1:N.
Before deployment, we also recommend conducting a POC under real-world conditions in your target environment, such as different lighting conditions, camera angles, face sizes, and mask usage, to validate recognition performance for your specific application.
FaceMe incorporates fairness and performance across demographic groups as key considerations in model development and evaluation. Through continuous model optimization and testing, we work to minimize performance differences across demographic groups.
In real-world deployments, recognition accuracy can also be affected by factors such as image quality, lighting, camera angles, and operating conditions. The U.S. National Institute of Standards and Technology (NIST) FRTE includes Demographic Effects evaluations that analyze facial recognition performance across age, gender, and demographic groups, providing an independent reference for assessing algorithm performance.
Before deployment, we also recommend conducting a POC using samples representative of the target user population and actual operating environment to validate FaceMe performance for the intended use case.
Yes. FaceMe offers recognition models with different accuracy levels and an adjustable matching threshold, allowing customers to achieve an appropriate balance between the False Acceptance Rate (FAR) and False Rejection Rate (FRR) based on their security requirements and use case.
For example, high-security applications such as access control or financial identity verification may use a stricter matching threshold, while convenience-oriented applications such as time and attendance can use different settings based on operational requirements.
We recommend testing under actual operating conditions during the POC stage to determine the appropriate recognition model and matching threshold.
1:1 face verification compares a captured face against the face template of a specific identity to confirm whether the person is who they claim to be. It is commonly used for digital identity verification (eKYC), secure login, and authentication.
1:N face identification searches a captured face against multiple face templates in a database to determine who the person is or whether they are present in the database. It is commonly used for blocklist detection, VIP recognition, and face search applications. Because 1:N typically requires a larger face database to be maintained over time, it also requires appropriate security measures for data encryption, access control, and database management.
Both FaceMe SDK and FaceMe Platform support 1:1 face verification and 1:N face identification, allowing customers to select the appropriate product based on their application and system architecture. Simply put, use 1:1 to verify a specific identity and 1:N to search for a possible identity within a database.
FaceMe eKYC combines facial recognition, liveness detection, and AI-powered recognition and OCR for Taiwan identity documents, including National ID cards, National Health Insurance (NHI) cards, driver’s licenses, vehicle registration certificates, Alien Resident Certificates (ARC), and passports. For global applications, it also supports e-Passport verification, helping financial institutions build more secure and seamless digital identity verification (eKYC) processes.
Applications include online account opening, credit applications, insurance enrollment, mobile banking login, and transaction authentication, helping streamline identity verification, reduce manual processing, and lower the risk of identity fraud.
FaceMe has been adopted by financial institutions and insurers including Chunghwa Post, Yuanta Bank, Yuanta Life, Capital Securities, and Generali Vietnam, supporting digital identity verification and facial recognition across a range of financial services and insurance applications.
FaceMe can support retail security, VIP customer experiences, and self-service applications. System integrators can deploy FaceMe Security as a turnkey solution, integrating with existing IP cameras and VMS platforms to quickly enable facial recognition applications. For example, retailers can detect individuals on a watchlist, such as known shoplifters, and alert store personnel in real time, or recognize enrolled VIP customers to provide more timely and personalized services.
FaceMe can also be integrated into self-checkout, member identification, and identity verification processes, helping retailers strengthen security while improving the customer experience and operational efficiency.
FaceMe can support a wide range of identity authentication and security applications for government and public-sector organizations, including two-factor authentication (2FA) for PC login, employee access control and time and attendance, visitor access management, and watchlist detection.
For security monitoring, FaceMe Security can integrate with existing IP cameras and VMS platforms to detect individuals on a watchlist and provide real-time alerts.
FaceMe is commonly used in manufacturing facilities for site access control, workforce security, time and attendance, buddy-punching prevention, visitor management, and contractor management, helping manufacturers improve the security and efficiency of workforce access management.
Facial recognition can also be integrated into production and quality-control workflows for operator identification and traceability. For example, in vehicle inspection processes, inspection records can be linked to the identity of the responsible operator, helping improve accountability and operational traceability.
FaceMe can be deployed across airports, transit stations, transportation hubs, and buses. By integrating FaceMe Security with existing IP cameras and VMS platforms, transportation operators can enable real-time watchlist detection and alerts, helping security teams quickly identify where persons of interest are detected and respond accordingly.
For cross-camera person search and tracking, People Tracker can help locate and follow specific individuals across different cameras. Even when a clear facial image is unavailable, it can use body and appearance attributes, such as clothing and body characteristics, for Person Re-Identification (Re-ID), helping security teams trace a person's movement across cameras and locations.
The system first creates a face database of enrolled individuals. Faces captured by cameras can then be recognized in real time using FaceMe Security or through integration with FaceMe SDK / FaceMe Platform APIs.
For access control, 1:1 face verification or 1:N face identification can be used to verify an individual’s identity and authorize access through the access control system. For security surveillance, faces can be matched against a watchlist, with real-time alerts generated when a person of interest is detected.
FaceMe enables cardless and passwordless facial recognition for employee clock-in and clock-out, helping verify employee identity and reduce buddy punching.
Organizations can use the time and attendance features in FaceMe Security or integrate facial recognition into existing HR or attendance systems using FaceMe SDK or FaceMe Platform, depending on their existing system architecture.
FaceMe combines facial recognition and liveness detection to verify a user’s identity, providing a biometric authentication factor for two-factor or multi-factor authentication (2FA/MFA).
Using FaceMe Platform APIs or FaceMe SDK, organizations can combine facial authentication with passwords, device authentication, SMS OTPs, or other authentication mechanisms for applications such as online banking, app login, and high-risk transaction verification.
FaceMe supports both face search and Person Re-Identification (Re-ID). When a clear facial image is available, 1:N face identification can be used to search for a specific person. Even when a clear face is unavailable, Re-ID can use appearance attributes such as clothing and body characteristics to help search for and track a person across different cameras and locations.
Organizations can integrate face search and Re-ID capabilities into their existing systems using FaceMe SDK or FaceMe Platform APIs. For faster deployment with less development and integration effort, organizations can use FaceMe Security with the People Tracker Add-On for cross-camera person search and tracking.
If facial recognition needs to run directly on a PC, mobile device, POS system, facial recognition terminal, smart camera, kiosk, or other edge device, FaceMe SDK is recommended. It supports multiple operating systems and AI computing platforms, enabling face detection, feature extraction, and matching directly on the device for low-latency, offline, and edge AI applications.
If the edge device is primarily used to capture images and send them to a centralized server for facial recognition, FaceMe Platform API is recommended. FaceMe Platform provides server-based facial recognition through REST APIs, with centralized management, scalability, and load balancing for multi-device and large-scale enterprise deployments
FaceMe SDK supports Windows, Linux, Android, and iOS and is optimized for major edge-computing platforms including Intel, NVIDIA, Qualcomm, MediaTek, NXP, and Ambarella.
Supported platforms and configurations can vary by SDK version, so the exact requirements should be confirmed during technical evaluation.
Yes.
FaceMe provides technical evaluation resources, including a trial SDK, sample code, and integration documentation, to help development teams confirm compatibility and integration approach before purchasing.
Contact the sales team through the website and describe your technical evaluation needs to get started.
Depending on the product, integration scope, and the customer’s development capabilities, an initial deployment or POC can typically take anywhere from a few days to a few weeks.
FaceMe Security is a ready-to-deploy turnkey solution that typically requires no custom coding and can be deployed after system configuration and integration. FaceMe Platform provides REST APIs for faster integration with web, server-based, and enterprise applications. FaceMe SDK is designed for developers who need deeper integration into native applications, edge devices, or AIoT products, offering greater flexibility while requiring more software development expertise.
The timeline for full production deployment will vary depending on deployment scale, existing system integration, and testing requirements.
A POC is recommended to validate the integration approach and establish a realistic production timeline.
No. FaceMe provides pretrained facial recognition models, so customers do not need to prepare their own training data or train a model from scratch.
For deployments with challenging conditions, such as extreme lighting, unusual camera angles, or specialized operating environments, we recommend conducting a POC with representative real-world samples to validate recognition performance and optimize system settings before deployment.
Yes. FaceMe Security can integrate with existing VMS platforms, IP cameras, and access control systems, allowing organizations to add facial recognition to their existing security infrastructure without replacing their current equipment.
FaceMe Security supports leading VMS platforms, including Genetec Security Center, Milestone XProtect, AXIS Camera Station, Network Optix Nx Witness, Hanwha Wisenet WAVE, and VIVOTEK VSS. Facial recognition, real-time alerts, and event search capabilities can be integrated into existing video surveillance workflows, with available features varying by VMS platform.
Yes. FaceMe supports offline operation and on-premises deployment. Depending on the product and system architecture, face detection, feature extraction, and matching can be performed directly on edge devices or on-premises servers without an internet connection.
For example, FaceMe SDK can run facial recognition directly on edge devices, while FaceMe Platform, FaceMe eKYC, and FaceMe Security can be deployed within an organization's own environment. Network connectivity can be configured as needed for cross-site data synchronization, remote management, or other cloud-based services.
Yes. When upgrading the FaceMe facial recognition engine, forward compatibility of existing face templates is taken into consideration so that existing data can continue to be used.
If a new version involves major changes to the algorithm or model architecture, face templates may need to be updated or migrated. We recommend confirming version compatibility and migration requirements before upgrading to ensure a smooth transition of the existing face database.
Face size in an image can affect detection and recognition performance. FaceMe uses face size and Inter-Ocular Distance (IOD) as key indicators for evaluating whether a face image is suitable for recognition. The face should occupy at least 2.5% of the image, with an IOD of at least 32 pixels. For close-range applications such as access control, a face size of 8% of the image is recommended for better recognition performance.
Actual requirements may vary depending on camera resolution, capture distance, face angle, and use case.
FaceMe supports multi-face detection and recognition within a single frame. The actual number of faces that can be processed simultaneously depends on factors such as image resolution, face size and density, occlusion, and processing hardware performance.
For high-traffic environments such as airports and transit stations, we recommend conducting a POC based on the actual camera setup and hardware configuration to validate the required real-time detection and recognition performance.
FaceMe provides multiple layers of protection against facial spoofing attacks. Liveness Detection (PAD) helps determine whether a real person is present in front of the camera, protecting against presentation attacks such as printed photos and video replays displayed on a screen.
For AI-generated or manipulated facial content, FaceMe also provides Deepfake Detection. In digital identity verification workflows, Injection Attack Detection (IAD) can provide an additional layer of protection by detecting attempts to bypass the camera and inject manipulated media directly into the image pipeline.
Together, Liveness Detection, Deepfake Detection, and IAD provide complementary layers of protection against both physical presentation attacks and digital injection attacks.
FaceMe converts facial images into irreversible face templates, which are mathematical representations used for face matching and search, rather than directly storing the original facial images. The original facial image cannot be reconstructed from the face template.
For data security, FaceMe uses AES-256 encryption to protect stored data and supports HTTPS / TLS 1.3 to protect data in transit. Data storage and retention policies can be managed according to an organization’s system architecture, security policies, and regulatory requirements.
FaceMe incorporates privacy protection and data governance mechanisms to help organizations address GDPR and other applicable local or regional privacy requirements. Depending on the product and deployment architecture, FaceMe supports on-premises or private-cloud deployment, data encryption, access controls, and configurable data retention policies, giving organizations greater control over facial and biometric data.
For example, FaceMe Security allows organizations to configure data retention periods based on regional regulations, use cases, and internal policies, and can automatically delete or anonymize personal data according to the defined retention schedule.
FaceMe is also certified to ISO/IEC 27001 for information security management and ISO/IEC 27701 for privacy information management.
Actual compliance requirements depend on the deployment location, data processing practices, and specific use case, and organizations should configure their deployment and data governance policies according to applicable regulations.
FaceMe offers perpetual and subscription-based licensing options, depending on the product and deployment requirements. Licensing models and pricing metrics vary based on the product, deployment scale, and use case.
For recommended licensing options and pricing, please contact the FaceMe sales team for an assessment based on your specific requirements.
FaceMe offers free trial or evaluation licenses, with availability and evaluation periods varying by product. Organizations can start with a trial or Proof of Concept (POC) to evaluate recognition performance, hardware compatibility, and integration feasibility in their intended environment before moving to production deployment and purchase.
To request a trial, start a POC, or obtain pricing, submit your requirements through the Contact Us form, including the product of interest, use case, deployment environment, and estimated scale. The FaceMe sales team will help determine the appropriate product and deployment approach.
Book a Demo
