The JPEG XL Image Coding System: History, Features, Coding Tools, Design Rationale, and Future

📅 2025-06-06
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
To address the low compression efficiency and functional fragmentation of legacy image formats (e.g., JPEG, PNG, GIF), this paper introduces JPEG XL—a unified, general-purpose next-generation image coding standard. JPEG XL employs a single codec architecture to jointly support lossy and lossless compression, progressive rendering, animation, and responsive imaging—resolving longstanding format silos. Key technical innovations include: (1) the first practical lossless transcoding of JPEG images; (2) adaptive quantization; (3) modular entropy coding; (4) neural-network-enhanced preprocessing; and (5) reversible color transformations—all designed to maximize compression performance while maintaining low decoding complexity. Experimental results demonstrate that JPEG XL achieves, on average, 60% bit-rate reduction over JPEG at equivalent subjective quality, and outperforms PNG in lossless compression efficiency. JPEG XL has been formally standardized as ISO/IEC 18181.

Technology Category

Application Category

📝 Abstract
JPEG XL is a new image coding system offering state-of-the-art compression performance, lossless JPEG recompression, and advanced features. It aims to replace JPEG, PNG, GIF, and other formats with a single universal codec. This article provides an overview of JPEG XL, including its history, design rationale, coding tools, and future potential. It can be used as a companion document to the standard (ISO/IEC 18181), or as a standalone article to better understand JPEG XL, either at a high level or in considerable technical detail.
Problem

Research questions and friction points this paper is trying to address.

JPEG XL aims to replace multiple image formats with one universal codec.
It offers advanced compression and lossless JPEG recompression capabilities.
The paper explains JPEG XL's design, tools, and future potential.
Innovation

Methods, ideas, or system contributions that make the work stand out.

State-of-the-art image compression performance
Lossless JPEG recompression capability
Universal codec replacing multiple formats
🔎 Similar Papers
No similar papers found.
Jon Sneyers
Jon Sneyers
Cloudinary Research
image compressionlogic programming
Jyrki Alakuijala
Jyrki Alakuijala
Software Engineer at Google
Data compressionhuman vision
Luca Versari
Luca Versari
Google Research
compression
Z
Zolt'an Szabadka
Google Research, Zürich, Switzerland
S
Sami Boukortt
Google Research, Zürich, Switzerland
A
Amnon Cohen-Tidhar
Cloudinary Research, Petah Tikva, Israel
M
Moritz Firsching
Google Research, Zürich, Switzerland
E
Evgenii Kliuchnikov
Google Research, Zürich, Switzerland
T
Tal Lev-Ami
Cloudinary Research, Petah Tikva, Israel
E
Eric Portis
Cloudinary Research, Petah Tikva, Israel
Thomas Richter
Thomas Richter
Moving Picture Technologies, Fraunhofer IIS, Erlangen, Germany
Osamu Watanabe
Osamu Watanabe
Tokyo Institute of Technology
Theoretical Computer Science