GigE Vision, USB3 Vision, CoaXPress and more now benefit from the full power of JPEG XS
JPEG XS is now officially supported as a compression technology within the GenICam framework, managed by The European Machine Vision Association (EMVA). The addition was confirmed in EMVA's GenICam Working Group Update, released following the International Vision Standard Meeting (IVSM) held in Prague in April 2026.
This milestone represents a major advancement for the machine vision industry, enabling higher performance, improved efficiency, and greater interoperability across imaging systems and devices.
What's new: SFNC 2.8 and GenDC 1.2
EMVA's update covers two related GenICam modules:
- Standard Features Naming Convention (SFNC) 2.8 — adds JPEG XS as a new supported compression/encoding option, alongside a range of other feature and unit updates.
- GenDC 1.2 — introduces native support for JPEG XS payloads, with additional clarifications to ensure smoother interoperability across implementations.
GenDC is the data-container specification within GenICam that provides a universal, interface-independent way to transport image and metadata streams from a wide range of devices — primarily cameras. With JPEG XS now integrated at this level, machine vision systems can move JPEG XS content through the same universal interface already used for raw and other formats.
Where this applies
Because GenDC sits underneath GenICam's various physical transports, JPEG XS support here flows straight through to the interfaces machine vision integrators already use day to day: USB3 Vision, GigE Vision, CoaXPress, and Camera Link. That means one standards-based integration path, rather than interface-specific workarounds, for bringing ultra-low-latency compressed video into existing camera ecosystems.
Why JPEG XS for machine vision
With JPEG XS now integrated, machine vision systems can fully leverage the benefits of ultra-low latency and lossless, near-lossless or visually lossless compression while maximizing compatibility across platforms.
JPEG XS also uniquely supports direct compression of raw Bayer CFA content, eliminating the need for debayering before encoding. This is particularly valuable for industrial and machine vision cameras, where preserving image fidelity, minimizing latency, and optimizing bandwidth are essential — especially as sensor resolutions and frame rates continue to climb.
intoPIX's role
As a key contributor to this evolution, intoPIX delivers advanced lightweight compression and image processing solutions for industrial and machine vision applications. intoPIX technologies enable visually lossless quality while supporting higher resolutions, lower latency, and increased transfer rates across existing interfaces.
intoPIX is proud to have contributed to this collaboration between JPEG XS and the GenICam standard, helping drive innovation, interoperability, and next-generation performance for machine vision systems.
Learn more: Read EMVA's full GenICam Working Group Update for the complete list of SFNC 2.8 and GenDC 1.2 changes.
Not yet familiar with JPEG XS?
JPEG XS is the first ISO JPEG Standard created for those who do not like compression.
JPEG XS offers extremely low latency (a few video lines only — less than 1 millisecond), extremely low complexity in hardware (FPGA, ASIC), extremely fast processing in software (CPU, GPU), and lossless quality (ISO JPEG calls it "transparent quality") for both human vision and machine learning / perception models.
Related content
intoPIX TicoXS is the new JPEG XS standard delivers the world smallest latency, best quality and lowest complexity in software and hardware.
intoPIX TicoXS is the new JPEG XS standard delivers the world smallest latency, best quality and lowest complexity in software and hardware.
Available for CPU or GPU, our TicoRAW SDKs provide the highest quality, visually lossless codec for the optimization of your application's infrastructure.
Available for CPU or GPU, our TicoRAW SDKs provide the highest quality, visually lossless codec for the optimization of your application's infrastructure.
