Deep learning image compression with multi-channel tANS coding and hardware deployment

Deep learning-based image compression outperforms traditional methods in coding efficiency, but its computational complexity hinders real-time deployment on embedded devices. This paper proposes a heterogeneous computing system combining GPU-accelerated inference and CPU-accelerated entropy coding v...

Full description

Saved in:
Bibliographic Details
Published in:Journal of real-time image processing Vol. 23; no. 1; p. 1
Main Authors: Zhu, Yaohua, Zhang, Yong, Liu, Ya, Jiang, Jingyu, Zhu, Yanghang, Huang, Mingsheng
Format: Journal Article
Language:English
Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.01.2026
Springer Nature B.V
Subjects:
Algorithms
Co-design
Coding
Compression ratio
Computer Graphics
Computer Science
Deep learning
Efficiency
Embedded systems
Entropy
Graphics processing units
Hardware
Image compression
Image Processing and Computer Vision
Image quality
Lookup tables
Multimedia Information Systems
Neural networks
Pattern Recognition
Real time
Signal,Image and Speech Processing
Wavelet transforms
Entropy coding
Deep learning image compression
tANS coding
Cache-friendly
GPU
ISSN:1861-8200, 1861-8219
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Deep learning-based image compression outperforms traditional methods in coding efficiency, but its computational complexity hinders real-time deployment on embedded devices. This paper proposes a heterogeneous computing system combining GPU-accelerated inference and CPU-accelerated entropy coding via lookup tables, breaking performance bottlenecks through algorithm-hardware co-design. After GPU acceleration, entropy coding becomes the dominant bottleneck (73% of runtime). To address this, we introduce three key innovations: replacing rANS with tANS encoding, converting dynamic computations into static table lookups, reducing encoding latency; a cache-friendly tANS coding scheme for the 192-channel network outputs, minimizing access latency; an out-of-range symbol encoding method, ensuring lossless and efficient compression. Experiments demonstrate that under high compression ratios, compared with traditional rANS, tANS reduces latency by 77%, with a compression ratio loss of 12.6% while still ensuring image compression quality higher than JPEG2000.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1861-8200
1861-8219
DOI:10.1007/s11554-025-01795-8