ConvNeXt-ChARM: ConvNeXt-based Transform for Efficient Neural Image Compression - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

ConvNeXt-ChARM: ConvNeXt-based Transform for Efficient Neural Image Compression

Résumé

In recent years, neural image compression has garnered considerable attention from both research and industry. It has shown great promise in surpassing traditional methods in terms of rate-distortion performance through the development of end-to-end deep neural codecs. Despite these advancements, there is still room for improvement, particularly in reducing the coding rate while maintaining high reconstruction fidelity, especially in non-homogeneous textured image areas. Current models, including attention-based transform coding, also tend to have a higher number of parameters and longer decoding times. To address these challenges, we propose ConvNeXt-ChARM, an efficient ConvNeXt-based transform coding framework. It is coupled with a compute-efficient channel-wise auto-regressive prior that captures both global and local contexts from the hyper and quantized latent representations. Our architecture can be optimized end-to-end, fully leveraging context information to extract compact latent representations and achieve higher-quality image reconstructions. Experimental results conducted on four widely-used datasets demonstrate the effectiveness of ConvNeXt-ChARM. It consistently delivers significant BD-rate (PSNR) reductions, averaging 5.24% over the VVC reference encoder (VTM-18.0) and 1.22% over the state-of-the-art learned image compression method SwinT-ChARM. Additionally, we conduct model scaling studies to verify the computational efficiency of our approach. Furthermore, we perform objective and subjective analyses to highlight the performance gap between ConvNeXt, the next-generation ConvNet, and the Swin Transformer. Overall, our proposed ConvNeXt-ChARM framework showcases improved compression efficiency and reconstruction quality, establishing itself as a promising solution in the field of neural image compression.

Dates et versions

hal-04356730 , version 1 (20-12-2023)

Identifiants

Citer

Ahmed Ghorbel, Wassim Hamidouche, Luce Morin. ConvNeXt-ChARM: ConvNeXt-based Transform for Efficient Neural Image Compression. 2023 11th European Workshop on Visual Information Processing (EUVIP), Sep 2023, Gjovik, Norway. ⟨10.1109/euvip58404.2023.10323063⟩. ⟨hal-04356730⟩
16 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More