Bi-modal palm recognition technology is one of the safest and fastest ways to verify the identity of a user through an access control terminal. It is no surprise that the technology is gaining traction in application scenarios with the highest standards.
Using palm features as an identifier is nothing new, but unlocking the full potential of the “information” embedded in the human hand requires a lot of technological finesse. Earlier generations of palm readers, for example, relied solely on capturing surface features, similar to fingerprint readers. As the area of the palm is larger, they captured more information than the former, but they did not add an additional layer of security. They simply captured more information of the same kind.
Moreover, they required precise deployment—users had to place their palm on the reader in the same position as they did when their user profile was created and their features entered the database. Like fingerprint sensors, these “old-generation” readers were also prone to error when the palmprint of the user showed signs of wear, as is common over people’s lifetime, especially in demanding workplace scenarios.
Armatura’s ARMLivePalm 9.6 SDK Bi-Modal Palm Recognition Technology represents a major leap forward as it combines visible-light palm shape analysis with near-infrared palm vein mapping. It is more accurate than previous generations of readers and better at detecting presentation attacks, thanks to using two modes of identity verification simultaneously. It is also faster and more stable in varying environmental conditions, which can unlock new use cases for the technology.
The human palm provides two independent layers of biometric data. Externally, there are the palm’s visible geometry, lines, texture, curvature and finger spacing. Internally, there are subcutaneous vein patterns formed by the body’s vascular structure. Vein patterns are unique even between identical twins and remain stable over time because they’re protected from wear or surface damage.
This dual information makes the palm one of the richest biometric surfaces available — more (and more diverse) data points than fingerprints, and more robust to aging or surface changes.
Cutting-edge performance engineered in the US
Armatura’s ARMLivePalm 9.6 SDK Bi-Modal Palm Recognition Technology is AN algorithmic platform that can be integrated in a wide array of access control terminals and gates, as well as mobile devices. Engineered in the US, it is aimed at professional-grade security deployments.
ARMLivePalm 9.6 can perform full authentication cycles in under half a second. It supports databases of up to 100,000 users and maintains an exceptionally low false rejection rate—below 0.01 percent at a strict false acceptance rate of 0.001 percent. Most importantly, it sustains this level of precision across varied lighting conditions and hand positions.
Enabling advanced hand pose tolerance was a central requirement when Armatura developed ARMLivePalm 9.6. The goal was to enable users to move their palm naturally over the reader, instead of having to align it rigidly with it.
These advancements set ARMLivePalm 9.6 apart from earlier palm or fingerprint systems, which often forced a trade-off between speed, capacity, and security. Armatura’s latest technology shows that large-scale, touchless palm recognition can now combine enterprise-level accuracy with the convenience users expect in demanding environments such as corporate campuses, hospitals, or public venues.
Putting AI into practice
At the core of ARMLivePalm 9.6 is an advanced AI backbone that allows the system to capture and interpret the palm’s visible and invisible features in real time. The technology combines two powerful concepts of advanced machine learning: transformer-based architecture and GAN-driven data simulation.
Together, they enable the platform to operate more efficiently and generalize across diverse environments, while maintaining high accuracy even in challenging conditions such as low light, motion or partial occlusion.
Transformers are a type of neural network originally developed for natural language processing. Unlike earlier convolutional AI models, which process visual data pixel by pixel, transformers focus on distinct structures and relationships between different elements of an image, regardless of whether they are close to or distant from each other. The logic of transformers is similar to the way humans understand context in a situation and are therefore able to understand connections between different events.
Powered by transformer-based AI, ARMLivePalm 9.6 can analyze visible-light and near-infrared palm data simultaneously, identifying which features—texture, curvature, or vein patterns—carry the most reliable information and correlating them to achieve highest accuracy.
The training of the ARMLivePalm 9.6 algorithm involves Generative Adversarial Networks. GANs are characterized by involving two competing neural networks—a generator that creates synthetic images and a discriminator that learns to tell real from fake. This doesn’t just increase accuracy, but it also raises the system’s tolerance to legitimate differences in image data. Enabled by GAN training, ARMLivePalm 9.6 takes into account that the user’s skin tone might change over time, that they might have surface injuries or skin conditions, as well as a wide range of lighting conditions, ensuring that they don’t affect the system’s accuracy.
Ease of deployment
ARMLivePalm 9.6 was designed to adapt flexibly to different operational environments. It can serve as the primary mode of authentication in high-security or privacy-sensitive scenarios—such as laboratories, data centers or hospitals—where contactless identification and spoofing resistance are essential. It can also function as a supporting layer in hybrid systems, where it can be combined with other biometric methods such as facial recognition. This is especially relevant in scenarios where convenience and redundancy are key, or when system operators want to benefit from the extra layer of security without requiring a full overhaul of their infrastructure and database.
ARMLivePalm 9.6 was designed for easy integration across multiple platforms. It comes as an SDK running on Windows, Android, and embedded Linux systems. With a model footprint of less than 100mb, it can be deployed on a wide range of hardware, from full-scale terminals to compact edge devices.
For system integrators, this cross-platform compatibility helps reduce deployment costs by up to 40%, as ARMLivePalm 9.6 unlocks palm recognition without the need for proprietary readers.
Scalability was another focus area during its development. With APIs for C/C++, Java, and C#, it is suitable for systems with a single point of entry, as it is for complex multi-site networks. Built-in diagnostic and sandbox tools allow integrators to test updates and monitor performance in real time, regardless of size and characteristics of their architecture.
Beyond access control
ARMLivePalm 9.6 was designed with much more in mind than verifying identities and allowing or denying access. Thanks to open APIs, integrators can connect the technology with other domains such as time attendance, visitor management, and transaction authorization. It can synchronize data with those systems, streamlining operations in high-demand environments from corporate offices and hospitals to factories and transportation hubs.
In environments where privacy is key and facial recognition is therefore unsuitable or restricted, for example by the GDPR in Europe, ARMLivePalm 9.6 provides organizations with biologically unique user data that is not prone to misuse, for example the creation of deepfakes. Overall, the platform is designed with security in mind: biometric data are encrypted end to end, while the system’s over-the-air update framework ensures that performance and anti-spoofing algorithms remain current against evolving threats.
The platform's security is rigorously validated, as it meets the ISO/IEC 30107-3 Level 2 standards for presentation attack detection, with a false rejection rate of 1.0–1.5%.
Thanks to the richness of data generated by ARMLivePalm 9.6, its ease of use and security features, Armatura envisions that bi-modAl palm recognition will serve as a key node in cognitive ecosystems of the future. With ARMLivePalm 9.6 new possibilities open up for organizations. From access control terminals to connected workplaces and smart building infrastructures, ARMLivePalm 9.6 bridges physical identity and digital intelligence in ways that redefine what access control is capable of.
For more information about Armatura, please visit
https://www.armatura.us.