# An Unsupervised Information-Theoretic Perceptual Quality Metric

arXiv: https://arxiv.org/abs/2006.06752 · HTML: https://mlsanity.com/paper/2006.06752/

## Key facts

- Cited by 28 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Jun 2020.

## Built on by (most cited first, 28 of 28)

- [Improving Statistical Fidelity for Neural Image Compression with Implicit Local Likelihood Models](https://mlsanity.com/paper/2301.11189/) (2023)
- [An Introduction to Neural Data Compression](https://mlsanity.com/paper/2202.06533/) (2022)
- [DICE: Diversity in Deep Ensembles via Conditional Redundancy Adversarial Estimation](https://mlsanity.com/paper/2101.05544/) (2021)
- [Shift-tolerant Perceptual Similarity Metric](https://mlsanity.com/paper/2207.13686/) (2022)
- [A Perceptual Quality Metric for Video Frame Interpolation](https://mlsanity.com/paper/2210.01879/) (2022)
- [Multiscale Sliced Wasserstein Distances as Perceptual Color Difference Measures](https://mlsanity.com/paper/2407.10181/) (2024)
- [Good, Cheap, and Fast: Overfitted Image Compression with Wasserstein Distortion](https://mlsanity.com/paper/2412.00505/) (2024)
- [On the relation between statistical learning and perceptual distances](https://mlsanity.com/paper/2106.04427/) (2021)
- [DeepDC: Deep Distance Correlation as a Perceptual Image Quality Evaluator](https://mlsanity.com/paper/2211.04927/) (2022)
- [Attacking Perceptual Similarity Metrics](https://mlsanity.com/paper/2305.08840/) (2023)
- [Approaching Peak Ground Truth](https://mlsanity.com/paper/2301.00243/) (2022)
- [ChatVis: Large Language Model Agent for Generating Scientific Visualizations](https://mlsanity.com/paper/2507.23096/) (2025)
- [Identifying and Mitigating Flaws of Deep Perceptual Similarity Metrics](https://mlsanity.com/paper/2207.02512/) (2022)
- [Video Coding Using Learned Latent GAN Compression](https://mlsanity.com/paper/2207.04324/) (2022)
- [Human Imperceptible Attacks and Applications to Improve Fairness](https://mlsanity.com/paper/2111.15603/) (2021)
- [The Unreasonable Effectiveness of Linear Prediction as a Perceptual Metric](https://mlsanity.com/paper/2310.05986/) (2023)
- [A Taxonomy of Miscompressions: Preparing Image Forensics for Neural Compression](https://mlsanity.com/paper/2409.05490/) (2024)
- [Content-Diverse Comparisons improve IQA](https://mlsanity.com/paper/2211.05215/) (2022)
- [Assessing invariance to affine transformations in image quality metrics](https://mlsanity.com/paper/2407.17927/) (2024)
- [Curiously Effective Features for Image Quality Prediction](https://mlsanity.com/paper/2106.05946/) (2021)
- [Image Statistics Predict the Sensitivity of Perceptual Quality Metrics](https://mlsanity.com/paper/2303.09874/) (2023)
- [CPIPS: Learning to Preserve Perceptual Distances in End-to-End Image Compression](https://mlsanity.com/paper/2310.00559/) (2023)
- [Quality assessment of 3D human animation: Subjective and objective evaluation](https://mlsanity.com/paper/2505.23301/) (2025)
- [Non-Aligned Reference Image Quality Assessment for Novel View Synthesis](https://mlsanity.com/paper/2511.08155/) (2025)
- [Steerable Pyramid Transform Enables Robust Left Ventricle Quantification](https://mlsanity.com/paper/2201.08388/) (2022)
- [Exploring Compressed Image Representation as a Perceptual Proxy: A Study](https://mlsanity.com/paper/2401.07200/) (2024)
- [Evaluating Perceptual Distance Models by Fitting Binomial Distributions to Two-Alternative Forced Choice Data](https://mlsanity.com/paper/2403.10390/) (2024)
- [Structured Uncertainty Similarity Score (SUSS): Learning a Probabilistic, Interpretable, Perceptual Metric Between Images](https://mlsanity.com/paper/2512.03701/) (2025)

## Abstract (arXiv)

Tractable models of human perception have proved to be challenging to build. Hand-designed models such as MS-SSIM remain popular predictors of human image quality judgements due to their simplicity and speed. Recent modern deep learning approaches can perform better, but they rely on supervised data which can be costly to gather: large sets of class labels such as ImageNet, image quality ratings, or both. We combine recent advances in information-theoretic objective functions with a computational architecture informed by the physiology of the human visual system and unsupervised training on pairs of video frames, yielding our Perceptual Information Metric (PIM). We show that PIM is competitive with supervised metrics on the recent and challenging BAPPS image quality assessment dataset and outperforms them in predicting the ranking of image compression methods in CLIC 2020. We also perform qualitative experiments using the ImageNet-C dataset, and establish that PIM is robust with respect to architectural details.