# Deep Perceptual Image Quality Assessment for Compression

arXiv: https://arxiv.org/abs/2103.01114 · HTML: https://mlsanity.com/paper/2103.01114/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 25 references that are in the map.
- First posted on arXiv 1 Mar 2021.

## Builds on (most cited first, 11 of 11)

- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Densely Connected Convolutional Networks](https://mlsanity.com/paper/1608.06993/) (2016)
- [EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks](https://mlsanity.com/paper/1905.11946/) (2019)
- [Identity Mappings in Deep Residual Networks](https://mlsanity.com/paper/1603.05027/) (2016)
- [NIMA: Neural Image Assessment](https://mlsanity.com/paper/1709.05424/) (2017)
- [KonIQ-10k: An ecologically valid database for deep learning of blind image quality assessment](https://mlsanity.com/paper/1910.06180/) (2019)
- [Understanding SSIM](https://mlsanity.com/paper/2006.13846/) (2020)
- [The Rate-Distortion-Accuracy Tradeoff: JPEG Case Study](https://mlsanity.com/paper/2008.00605/) (2020)
- [Lossy Compression with Distortion Constrained Optimization](https://mlsanity.com/paper/2005.04064/) (2020)
- [Rank-smoothed Pairwise Learning In Perceptual Quality Assessment](https://mlsanity.com/paper/2011.10893/) (2020)
- [Perceptually Optimizing Deep Image Compression](https://mlsanity.com/paper/2007.02711/) (2020)

## Built on by (most cited first, 9 of 9)

- [Exponentiated Gradient Reweighting for Robust Training Under Label Noise and Beyond](https://mlsanity.com/paper/2104.01493/) (2021)
- [DriftRec: Adapting diffusion models to blind JPEG restoration](https://mlsanity.com/paper/2211.06757/) (2022)
- [MSTRIQ: No Reference Image Quality Assessment Based on Swin Transformer with Multi-Stage Fusion](https://mlsanity.com/paper/2205.10101/) (2022)
- [Blind Image Quality Assessment via Transformer Predicted Error Map and Perceptual Quality Token](https://mlsanity.com/paper/2305.09353/) (2023)
- [Stay Positive: Non-Negative Image Synthesis for Augmented Reality](https://mlsanity.com/paper/2202.00659/) (2022)
- [Foreground Focus: Enhancing Coherence and Fidelity in Camouflaged Image Generation](https://mlsanity.com/paper/2504.02180/) (2025)
- [UlRe-NeRF: 3D Ultrasound Imaging through Neural Rendering with Ultrasound Reflection Direction Parameterization](https://mlsanity.com/paper/2408.00860/) (2024)
- [Pixel-by-pixel Mean Opinion Score (pMOS) for No-Reference Image Quality Assessment](https://mlsanity.com/paper/2206.06541/) (2022)
- [Super Efficient Neural Network for Compression Artifacts Reduction and Super Resolution](https://mlsanity.com/paper/2401.14641/) (2024)

## Abstract (arXiv)

Lossy Image compression is necessary for efficient storage and transfer of data. Typically the trade-off between bit-rate and quality determines the optimal compression level. This makes the image quality metric an integral part of any imaging system. While the existing full-reference metrics such as PSNR and SSIM may be less sensitive to perceptual quality, the recently introduced learning methods may fail to generalize to unseen data. In this paper we propose the largest image compression quality dataset to date with human perceptual preferences, enabling the use of deep learning, and we develop a full reference perceptual quality assessment metric for lossy image compression that outperforms the existing state-of-the-art methods. We show that the proposed model can effectively learn from thousands of examples available in the new dataset, and consequently it generalizes better to other unseen datasets of human perceptual preference.