Selected work

Representative Papers

Estimating SSIM from MSE for DCT-Based Compressed Images

Aug 03, 2026

This work addresses the challenge of accurately estimating the perceptual quality of DCT-compressed images, such as JPEG, using only global mean squared error (MSE) and original image statistics. Existing approaches struggle to effectively approximate the Structural Similarity Index (SSIM) under these constraints. The paper proposes a novel method that redistributes global MSE into local MSE estimates by leveraging the reference image’s local variance or standard deviation, thereby enabling accurate SSIM approximation without requiring ground-truth local error maps. Grounded in the characteristics of DCT-domain compression, the approach naturally extends to video applications. Experimental results on the Kodak and Xiph Subset1 datasets demonstrate that the proposed method significantly outperforms the global MSE baseline across a wide range of JPEG quality levels, achieving both high accuracy and robustness in SSIM estimation.

0 citationsRead paper

UI-LIC: A Unified Framework for Evaluating Learned Image Compression Models

Jun 22, 2026

This work addresses the challenge of fair comparison in learned image compression (LIC), which has been hindered by inconsistent model implementations, training protocols, and evaluation metrics. To this end, we present UI-LIC, an open-source unified framework that integrates six state-of-the-art LIC models alongside traditional codecs within a consistent experimental setup, enabling end-to-end automated training, inference, and comparative evaluation. The framework features a graphical user interface supporting bitrate alignment, computation of multiple quality metrics—including PSNR, SSIM, VMAF, and LPIPS—and interactive visualization of quality heatmaps. Deployment and benchmarking require only a single command, substantially lowering the barrier to entry for researchers. The code is publicly released to foster reproducibility and further advancement in the field.

0 citationsRead paper

Recent publications

Latest Papers

Estimating SSIM from MSE for DCT-Based Compressed Images

Aug 03, 2026

This work addresses the challenge of accurately estimating the perceptual quality of DCT-compressed images, such as JPEG, using only global mean squared error (MSE) and original image statistics. Existing approaches struggle to effectively approximate the Structural Similarity Index (SSIM) under these constraints. The paper proposes a novel method that redistributes global MSE into local MSE estimates by leveraging the reference image’s local variance or standard deviation, thereby enabling accurate SSIM approximation without requiring ground-truth local error maps. Grounded in the characteristics of DCT-domain compression, the approach naturally extends to video applications. Experimental results on the Kodak and Xiph Subset1 datasets demonstrate that the proposed method significantly outperforms the global MSE baseline across a wide range of JPEG quality levels, achieving both high accuracy and robustness in SSIM estimation.

0 citationsRead paper

UI-LIC: A Unified Framework for Evaluating Learned Image Compression Models

Jun 22, 2026

This work addresses the challenge of fair comparison in learned image compression (LIC), which has been hindered by inconsistent model implementations, training protocols, and evaluation metrics. To this end, we present UI-LIC, an open-source unified framework that integrates six state-of-the-art LIC models alongside traditional codecs within a consistent experimental setup, enabling end-to-end automated training, inference, and comparative evaluation. The framework features a graphical user interface supporting bitrate alignment, computation of multiple quality metrics—including PSNR, SSIM, VMAF, and LPIPS—and interactive visualization of quality heatmaps. Deployment and benchmarking require only a single command, substantially lowering the barrier to entry for researchers. The code is publicly released to foster reproducibility and further advancement in the field.

0 citationsRead paper