mlsanityData to 14 Oct 2026

ProxIQA: A Proxy Approach to Perceptual Optimization of Learned Image Compression

arXiv 1910.08845 ↗First posted 19 Oct 2019eess.IV cs.CV cs.LGcited by 24

Not on the map. Why?

Abstract (arXiv)

The use of ℓp (p=1,2) norms has largely dominated the measurement of loss in neural networks due to their simplicity and analytical properties. However, when used to assess the loss of visual information, these simple norms are not very consistent with human perception. Here, we describe a different "proximal" approach to optimize image analysis networks against quantitative perceptual models. Specifically, we construct a proxy network, broadly termed ProxIQA, which mimics the perceptual model while serving as a loss layer of the network. We experimentally demonstrate how this optimization framework can be applied to train an end-to-end optimized image compression network. By building on top of an existing deep image compression model, we are able to demonstrate a bitrate reduction of as much as 31% over MSE optimization, given a specified perceptual quality (VMAF) level.

Key facts. Cited by 24 papers in the map (data through 14 Oct 2026). First posted on arXiv 19 Oct 2019.

Builds on

No reference list yet.

Built on by

Cited by year

Citing papers per year: 2021: 8; 2022: 8; 2023: 3; 2024: 5.

24 papers.

Sort:

  1. 2022MaxViT: Multi-Axis Vision Transformercited by 476
  2. 2022MAXIM: Multi-Axis MLP for Image Processingcited by 302
  3. 2021RAPIQUE: Rapid and Accurate Video Quality Prediction of User Generated Contentcited by 100
  4. 2024Video Quality Assessment: A Comprehensive Surveycited by 15
  5. 2023Robust Unsupervised StyleGAN Image Restorationcited by 14
  6. 2022FAVER: Blind Quality Prediction of Variable Frame Rate Videoscited by 10
  7. 2021Hacking VMAF and VMAF NEG: vulnerability to different preprocessing methodscited by 8
  8. 2022Pik-Fix: Restoring and Colorizing Old Photoscited by 8
  9. 2022Semantically Video Coding: Instill Static-Dynamic Clues into Structured Bitstream for AI Taskscited by 7
  10. 2023On the Computation of the Gaussian Rate-Distortion-Perception Functioncited by 5
  11. 2021Regression or Classification? New Methods to Evaluate No-Reference Picture and Video Quality Modelscited by 4
  12. 2021Learning-Based Video Coding with Joint Deep Compression and Enhancementcited by 4
12 more
  1. 2023VMAF Re-implementation on PyTorch: Some Experimental Resultscited by 4
  2. 2024Image Quality Assessment: Investigating Causal Perceptual Effects with Abductive Counterfactual Inferencecited by 4
  3. 2021Scale-Consistent Fusion: from Heterogeneous Local Sampling to Global Immersive Renderingcited by 2
  4. 2021HLIC: Harmonizing Optimization Metrics in Learned Image Compression by Reinforcement Learningcited by 2
  5. 2021Self-Supervised Learning of Perceptually Optimized Block Motion Estimates for Video Compressioncited by 2
  6. 2024ACC-ViT : Atrous Convolution's Comeback in Vision Transformerscited by 2
  7. 2022QMRNet: Quality Metric Regression for EO Image Quality Assessment and Super-Resolutioncited by 1
  8. 2022Applicability limitations of differentiable full-reference image-qualitycited by 1
  9. 2024Perceptual Learned Image Compression via End-to-End JND-Based Optimizationcited by 1
  10. 2024A Sharpness Based Loss Function for Removing Out-of-Focus Blurcited by 1
  11. 2021Predicting Eye Fixations Under Distortion Using Bayesian Observerscited by 0
  12. 2022ROMNet: Renovate the Old Memoriescited by 0