Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Image Quality Assessment With Compressed Sampling

Ronghua Liao, Chen Hui, Lang Yuan, Haiqi Zhu, Feng Jiang

TL;DR

This work proposes two networks for NR-IQA with Compressive Sampling (dubbed CL-IQA and CS-IQA), which outperform other methods on various datasets with less data usage.

Abstract

No-Reference Image Quality Assessment (NR-IQA) aims at estimating image quality in accordance with subjective human perception. However, most methods focus on exploring increasingly complex networks to improve the final performance,accompanied by limitations on input images. Especially when applied to high-resolution (HR) images, these methods offen have to adjust the size of original image to meet model input.To further alleviate the aforementioned issue, we propose two networks for NR-IQA with Compressive Sampling (dubbed CL-IQA and CS-IQA). They consist of four components: (1) The Compressed Sampling Module (CSM) to sample the image (2)The Adaptive Embedding Module (AEM). The measurements are embedded by AEM to extract high-level features. (3) The Vision Transformer and Scale Swin TranBlocksformer Moudle(SSTM) to extract deep features. (4) The Dual Branch (DB) to get final quality score. Experiments show that our proposed methods outperform other methods on various datasets with less data usage.

Image Quality Assessment With Compressed Sampling

TL;DR

This work proposes two networks for NR-IQA with Compressive Sampling (dubbed CL-IQA and CS-IQA), which outperform other methods on various datasets with less data usage.

Abstract

No-Reference Image Quality Assessment (NR-IQA) aims at estimating image quality in accordance with subjective human perception. However, most methods focus on exploring increasingly complex networks to improve the final performance,accompanied by limitations on input images. Especially when applied to high-resolution (HR) images, these methods offen have to adjust the size of original image to meet model input.To further alleviate the aforementioned issue, we propose two networks for NR-IQA with Compressive Sampling (dubbed CL-IQA and CS-IQA). They consist of four components: (1) The Compressed Sampling Module (CSM) to sample the image (2)The Adaptive Embedding Module (AEM). The measurements are embedded by AEM to extract high-level features. (3) The Vision Transformer and Scale Swin TranBlocksformer Moudle(SSTM) to extract deep features. (4) The Dual Branch (DB) to get final quality score. Experiments show that our proposed methods outperform other methods on various datasets with less data usage.
Paper Structure (12 sections, 7 equations, 5 figures, 3 tables)

This paper contains 12 sections, 7 equations, 5 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Proposed Approach
  3. Compressed Sampling Module
  4. Adaptive Embedding Module
  5. Vision Transformer
  6. Scale Swin TranBlocksformer Moudle
  7. Dual Branch
  8. Experiments
  9. Experimental Settings
  10. Comparison with State-of-the-Art Methods
  11. Visualization
  12. Ablation Study

Figures (5)

  • Figure 1: Random-crop removes pixels from every block despite the importance, leading to severe information loss.
  • Figure 2: The architecture of our proposed CS-IQA and CL-IQA.
  • Figure 3: Process of Flexible Sampling Module (AEM).
  • Figure 4: The dual-branch (DB) structure. Each branch contains two fully connected layers for the score and weight prediction.
  • Figure 5: The dual-branch (DB) structure. Each branch contains two fully connected layers for the score and weight prediction.