用线性注意力提升图像压缩效率，兼顾性能与低复杂度。
Linear Attention Modeling for Learned Image Compression
- 采用双向RWKV块实现紧凑特征提取
- 在Kodak等数据集上比VTM-9.1低15.26%以上率失真
- 适合追求高效压缩的工程应用
近年来，学习型图像压缩在编码效率上取得显著进展，主要依赖非线性神经网络变换和可学习熵建模。然而多数研究聚焦于强大主干网络，较少关注低复杂度设计。本文提出LALIC，一种用于学习型图像压缩的线性注意力建模方法。特别地，我们引入双向RWKV块，通过空间混洗与通道混洗模块实现更紧凑的特征提取，并采用基于卷积的全向移位模块适配二维潜在表示。此外，提出基于RWKV的空间-通道上下文模型（RWKV-SCCTX），利用双向RWKV有效建模邻近特征的相关性。据我们所知，这是首个将高效双向RWKV模型与线性注意力应用于学习型图像压缩的工作。实验表明，该方法在Kodak、CLIC和Tecnick数据集上分别以-15.26%、-15.41%、-17.63%的BD-rate优势超越VTM-9.1。代码已开源：https://github.com/sjtu-medialab/RwkvCompress。
原文摘要 · Abstract (English)
Recent years, learned image compression has made tremendous progress to achieve impressive coding efficiency. Its coding gain mainly comes from non-linear neural network-based transform and learnable entropy modeling. However, most studies focus on a strong backbone, and few studies consider a low complexity design. In this paper, we propose LALIC, a linear attention modeling for learned image compression. Specially, we propose to use Bi-RWKV blocks, by utilizing the Spatial Mix and Channel Mix modules to achieve more compact feature extraction, and apply the Conv based Omni-Shift module to adapt to two-dimensional latent representation. Furthermore, we propose a RWKV-based Spatial-Channel ConTeXt model (RWKV-SCCTX), that leverages the Bi-RWKV to modeling the correlation between neighboring features effectively. To our knowledge, our work is the first work to utilize efficient Bi-RWKV models with linear attention for learned image compression. Experimental results demonstrate that our method achieves competitive RD performances by outperforming VTM-9.1 by -15.26%, -15.41%, -17.63% in BD-rate on Kodak, CLIC and Tecnick datasets. The code is available at https://github.com/sjtu-medialab/RwkvCompress .
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。