# Rank-smoothed Pairwise Learning In Perceptual Quality Assessment

arXiv: https://arxiv.org/abs/2011.10893 · HTML: https://mlsanity.com/paper/2011.10893/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Nov 2020.

## Built on by (most cited first, 5 of 5)

- [Imagen Editor and EditBench: Advancing and Evaluating Text-Guided Image Inpainting](https://mlsanity.com/paper/2212.06909/) (2022)
- [Deep Perceptual Image Quality Assessment for Compression](https://mlsanity.com/paper/2103.01114/) (2021)
- [Artifact- and content-specific quality assessment for MRI with image rulers](https://mlsanity.com/paper/2111.03780/) (2021)
- [MR-IQA: A Unified Margin View of Regression and Ranking for Blind Image Quality Assessment](https://mlsanity.com/paper/2606.29760/) (2026)
- [FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance](https://mlsanity.com/paper/2609.25716/) (2026)

## Abstract (arXiv)

Conducting pairwise comparisons is a widely used approach in curating human perceptual preference data. Typically raters are instructed to make their choices according to a specific set of rules that address certain dimensions of image quality and aesthetics. The outcome of this process is a dataset of sampled image pairs with their associated empirical preference probabilities. Training a model on these pairwise preferences is a common deep learning approach. However, optimizing by gradient descent through mini-batch learning means that the "global" ranking of the images is not explicitly taken into account. In other words, each step of the gradient descent relies only on a limited number of pairwise comparisons. In this work, we demonstrate that regularizing the pairwise empirical probabilities with aggregated rankwise probabilities leads to a more reliable training loss. We show that training a deep image quality assessment model with our rank-smoothed loss consistently improves the accuracy of predicting human preferences.