Interleaved Block-based Learned Image Compression with Feature Enhancement and Quantization Error Compensation
This work addresses key bottlenecks in learned image compression for applications requiring high-quality image storage and transmission, though it is incremental as it builds on existing methods.
The paper tackles the challenges of obtaining compact latent representations and reducing quantization errors in learned image compression by proposing four modules for feature extraction, refinement, enhancement, and quantization error compensation. Experiments show that integrating these modules with Tiny-LIC outperforms existing methods and standards in PSNR and MS-SSIM on the Kodak and CLIC datasets.
In recent years, learned image compression (LIC) methods have achieved significant performance improvements. However, obtaining a more compact latent representation and reducing the impact of quantization errors remain key challenges in the field of LIC. To address these challenges, we propose a feature extraction module, a feature refinement module, and a feature enhancement module. Our feature extraction module shuffles the pixels in the image, splits the resulting image into sub-images, and extracts coarse features from the sub-images. Our feature refinement module stacks the coarse features and uses an attention refinement block composed of concatenated three-dimensional convolution residual blocks to learn more compact latent features by exploiting correlations across channels, within sub-images (intra-sub-image correlations), and across sub-images (inter-sub-image correlations). Our feature enhancement module reduces information loss in the decoded features following quantization. We also propose a quantization error compensation module that mitigates the quantization mismatch between training and testing. Our four modules can be readily integrated into state-of-the-art LIC methods. Experiments show that combining our modules with Tiny-LIC outperforms existing LIC methods and image compression standards in terms of peak signal-to-noise ratio (PSNR) and multi-scale structural similarity (MS-SSIM) on the Kodak dataset and the CLIC dataset.