High-Performance Open-Source Eye Tracking for AR/VR

AUSTIN, TX — October 5, 2025 — Almalence, a leader in computational imaging, has announced the release of a groundbreaking new eye tracking algorithm for augmented and virtual reality (AR/VR) devices. This innovative solution, now available as open-source software, is poised to make eye tracking accessible, affordable, and adaptable for the next generation of AR/VR hardware.

Eye tracking technology has the potential to revolutionize user interaction in AR and VR — enabling features such as foveated rendering, enhanced analytics, intuitive controls, lifelike avatars, and even optical aberrations and blur compensation. However, despite its promise, widespread adoption has been limited by the high costs imposed by proprietary solution vendors, along with challenges related to latency, hardware compatibility, and power consumption.

Almalence’s new algorithm addresses these barriers head-on:

• Ultra-Fast & Low-Latency: Optimized for real-time performance, the solution delivers rapid and reliable tracking — essential for immersive experiences.
• Hardware-Agnostic: Unlike existing alternatives, Almalence’s technology works with virtually any camera and light source placement, supporting the unique requirements of diverse headset designs.
• Advanced 3D Tracking: The algorithm provides the z-coordinate of the pupil — a feature rarely available in existing solutions and often priced at a premium by current vendors.
• Highly Efficient Implementation: The algorithm features an effective C implementation for generic CPU architectures and natively fits DSP architectures such as Qualcomm Hexagon, enabling extremely low latency and low power consumption.
• Open-Source for Implementation: The algorithm is released under the GPLv3 license, allowing AR/VR device makers to freely implement it in their designs.

“Our goal is to break down the technical and financial barriers that have held back eye tracking adoption in AR/VR,” said Eugene Panich, co-founder of Almalence. “By releasing this high-performance, hardware-agnostic solution as open-source, we empower device manufacturers to rapidly integrate our technology without up-front costs or commitments.”

Developers and manufacturers can access the algorithm now, test its performance with their own hardware, and begin integrating advanced eye tracking into their devices.

For more information, technical details, and to access the open-source release, visit https://github.com/almalence/eyetracking.