MLIC ++ : Linear Complexity Multi-Reference Entropy Modeling for Learned Image Compression
ACM Transactions on Multimedia Computing, Communications, and Applications · 31.05.2025 · 21 (5) · 1-25
DOI: 10.1145/3719011 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Multimedia Computing, Communications, and Applications, online erschienen am 22.05.2025. Verfasst von Wei Jiang und 4 weiteren Autor:innen. Die Arbeit wurde bisher 46-mal zitiert (Crossref) und nennt 73 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guangdong Provincial Key Laboratory of Ultra High Definition Immersive Media Technology, National Natural Science Foundation of China, Shenzhen Science and Technology Program-Shenzhen Cultivation of Excellent Scientific and Technological Innovation Talents project.
Abstract (Originalsprache)
The latent representation in learned image compression encompasses channel-wise, local spatial, and global spatial correlations, which are essential for the entropy model to capture for conditional entropy minimization. Efficiently capturing these contexts within a single entropy model, especially in high-resolution image coding, presents a challenge due to the computational complexity of existing global context modules. To address this challenge, we propose the Linear Complexity Multi-Reference Entropy Model (MEM \({}^{++}\) ). Specifically, the latent representation is partitioned into multiple slices. For channel-wise contexts, previously compressed slices serve as the context for compressing a particular slice. For local contexts, we introduce a shifted-window-based checkerboard attention module. This module ensures linear complexity without sacrificing performance. For global contexts, we propose a linear complexity attention mechanism. It captures global correlations by decomposing the softmax operation, enabling the implicit computation of attention maps from previously decoded slices. Using MEM++ as the entropy model, we develop the image compression method MLIC \({}^{++}\) . Extensive experimental results demonstrate that MLIC \({}^{++}\) achieves state-of-the-art performance, reducing BD-rate by \(13.39\%\) on the Kodak dataset compared to VTM-17.0 in Peak Signal-to-Noise Ratio (PSNR). Furthermore, MLIC \({}^{++}\) exhibits linear computational complexity and memory consumption with resolution, making it highly suitable for high-resolution image coding. Code and pre-trained models are available at https://github.com/JiangWeibeta/MLIC .
Förderung
- Guangdong Provincial Key Laboratory of Ultra High Definition Immersive Media Technology
- National Natural Science Foundation of China
- Shenzhen Science and Technology Program-Shenzhen Cultivation of Excellent Scientific and Technological Innovation Talents project
- Outstanding Talents Training Fund in Shenzhen
Zitieren
APA
Jiang, W., Yang, J., Zhai, Y., Gao, F., & Wang, R. (2025). MLIC ++ : Linear Complexity Multi-Reference Entropy Modeling for Learned Image Compression. ACM Transactions on Multimedia Computing, Communications, and Applications, 21(5), 1-25. https://doi.org/10.1145/3719011
Vancouver
Jiang W, Yang J, Zhai Y, Gao F, Wang R. MLIC ++ : Linear Complexity Multi-Reference Entropy Modeling for Learned Image Compression. ACM Transactions on Multimedia Computing, Communications, and Applications. 2025;21(5):1-25. doi:10.1145/3719011
BibTeX
@article{Jiang2025,
author = {Jiang, Wei and Yang, Jiayu and Zhai, Yongqi and Gao, Feng and Wang, Ronggang},
title = {MLIC ++ : Linear Complexity Multi-Reference Entropy Modeling for Learned Image Compression},
journal = {ACM Transactions on Multimedia Computing, Communications, and Applications},
year = {2025},
volume = {21},
number = {5},
pages = {1--25},
doi = {10.1145/3719011}
}
EUVIMED