Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.
Awesome Lists | Featured Topics | Projects
https://github.com/chaofengc/Awesome-Image-Quality-Assessment

A comprehensive collection of IQA papers
https://github.com/chaofengc/Awesome-Image-Quality-Assessment
List: Awesome-Image-Quality-Assessment
awesome awesome-list blind-image-quality-assessment image-quality-assessment iqa iqa-dataset
Last synced: about 1 month ago
JSON representation
A comprehensive collection of IQA papers
Host: GitHub
URL: https://github.com/chaofengc/Awesome-Image-Quality-Assessment
Owner: chaofengc
License: mit
Created: 2021-11-28T13:29:05.000Z (about 3 years ago)
Default Branch: main
Last Pushed: 2024-11-05T12:05:44.000Z (about 2 months ago)
Last Synced: 2024-11-05T13:23:20.543Z (about 2 months ago)
Topics: awesome, awesome-list, blind-image-quality-assessment, image-quality-assessment, iqa, iqa-dataset
Language: TeX
Homepage:
Size: 97.7 KB
Stars: 990
Watchers: 25
Forks: 65
Open Issues: 5
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project

awesoMRI-QC - awesome-image-quality-assessment
ultimate-awesome - Awesome-Image-Quality-Assessment - A comprehensive collection of IQA papers. (Other Lists / Monkey C Lists)
README

        # Awesome Image Quality Assessment (IQA)

A comprehensive collection of IQA papers, datasets and codes. We also provide PyTorch implementations of mainstream metrics in [IQA-PyTorch](https://github.com/chaofengc/IQA-PyTorch) 

[![toolbox](https://img.shields.io/badge/Toolbox-IQA--PyTorch-critical)](https://github.com/chaofengc/IQA-PyTorch) [![PyPI](https://img.shields.io/pypi/v/pyiqa)](https://pypi.org/project/pyiqa/) ![visitors](https://visitor-badge.laobi.icu/badge?page_id=chaofengc/Awesome-Image-Quality-Assessment) 

- [Awesome Image Quality Assessment (IQA)](#awesome-image-quality-assessment-iqa)

  - [Papers](#papers)

    - [Explainable IQA](#explainable-iqa)

    - [AIGC IQA](#aigc-iqa)

    - [No Reference (NR)](#no-reference-nr)

    - [Full Reference (FR)](#full-reference-fr)

    - [Image Aesthetic Assessment](#image-aesthetic-assessment)

    - [Others](#others)

  - [Datasets](#datasets)

    - [IQA datasets](#iqa-datasets)

    - [Perceptual similarity datasets](#perceptual-similarity-datasets)

  

Related Resources:

- [Awesome Image Aesthetic Assessment and Cropping](https://github.com/bcmi/Awesome-Aesthetic-Evaluation-and-Cropping). A curated list of resources including papers, datasets, and relevant links to aesthetic evaluation and cropping.

## Papers

### Unified IQA

> All IQA types unified in a single model

- `[ECCV 2024]` [PromptIQA: Boosting the Performance and Generalization for No-Reference Image Quality Assessment via Prompts](https://arxiv.org/abs/2403.04993), Chen et al. [Bibtex](./iqa_ref.bib#L931-L938)

- `[ICML 2024]` [Q-Align: Teaching LMMs for Visual Scoring via Discrete Text-Defined Levels](https://arxiv.org/abs/2312.17090), Wu et al. [Github](https://github.com/TianheWu/MLLMs-for-IQA) | [Bibtex](./iqa_ref.bib#L867-L874)

### Explainable IQA 

> Human readable IQA, mostly with large language models

- `[TPAMI 2024]` [Q-Bench+: A Benchmark for Multi-modal Foundation Models on Low-level Vision from Single Images to Pairs](https://arxiv.org/abs/2402.07116), Zhang et al. [Github](https://github.com/Q-Future/Q-Bench) | [Bibtex](./iqa_ref.bib#L1044-L1049)

- `[ACM MM2024]` [Q-Ground: Image Quality Grounding with Large Multi-modality Models](https://arxiv.org/abs/2407.17035), Chen et al. [Bibtex](./iqa_ref.bib#L1002-L1007) | [Github](https://github.com/Q-Future/Q-Ground)

- `[Arxiv 2024]` [VisualCritic: Making LMMs Perceive Visual Quality Like Humans](https://arxiv.org/abs/2403.12806), Huang et al. [Bibtex](./iqa_ref.bib#L940-L947)

- `[ECCV 2024]` [A Comprehensive Study of Multimodal Large Language Models for Image Quality Assessment](https://arxiv.org/abs/2403.10854), Wu et al. [Github]() | [Bibtex](./iqa_ref.bib#L922-L929)

- `[ECCV 2024]` [Towards Open-ended Visual Quality Comparison](https://arxiv.org/abs/2402.16641), Wu et al. [Github](https://github.com/Q-Future/Co-Instruct) | [Bibtex](./iqa_ref.bib#L906-L913)

- `[ECCV 2024]` [Depicting Beyond Scores: Advancing Image Quality Assessment through Multi-modal Language Models](https://arxiv.org/abs/2312.08962), You et al. [Project](https://depictqa.github.io/) | [Bibtex](./iqa_ref.bib#L885-L890)

- `[CVPR 2024]` [Q-Instruct: Improving Low-level Visual Abilities for Multi-modality Foundation Models](https://arxiv.org/abs/2311.06783), Wu et al. [Github](https://github.com/Q-Future/Q-Instruct) | [Bibtex](./iqa_ref.bib#L876-L883)

### AIGC IQA

- ✨`[CVPR 2024 (best paper)]` [Rich Human Feedback for Text-to-Image Generation](https://arxiv.org/abs/2312.10240), Liang et al. [Github](https://github.com/google-research-datasets/richhf-18k) | [Bibtex](./iqa_ref.bib#L1067-L1072)

- `[ICLR 2024]` [Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Vision](https://arxiv.org/abs/2309.14181), Wu et al. [Github](https://github.com/Q-Future/Q-Bench) | [Bibtex](./iqa_ref.bib#L833-L838)

- `[ICCV 2023]` [TIFA: Text-to-Image Faithfulness Evaluation with Question Answering](https://arxiv.org/abs/2303.11897), Hu et al. [Github](https://github.com/Yushi-Hu/tifa) | [Bibtex](./iqa_ref.bib#L840-L846) | [Project](https://tifa-benchmark.github.io/)

- `[NeurIPS 2023]` [ImageReward: Learning and Evaluating Human Preferences for Text-to-image Generation](https://arxiv.org/abs/2304.05977), Xu et al. [Github](https://github.com/THUDM/ImageReward) | [Bibtex](./iqa_ref.bib#L817-L822)

- `[ICCV2023]` [Better Aligning Text-to-Image Models with Human Preference](https://arxiv.org/abs/2303.14420), Wu et al. [Github](https://github.com/tgxs002/align_sd) | [Github(HPSv2)](https://github.com/tgxs002/HPSv2) | [Bibtex](./iqa_ref.bib#L804-L809)

- `[NeurIPS 2023]` [Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation](https://arxiv.org/abs/2305.01569), Yuval et al. [Github](https://github.com/yuvalkirstain/PickScore) | [Bibtex](./iqa_ref.bib#L811-L815)

- `[TCSVT2023]` [A Fine-grained Subjective Perception & Alignment Database for AI Generated Image Quality Assessment](https://arxiv.org/abs/2306.04717), Li et al. [Github](https://github.com/lcysyzxdxc/AGIQA-3k-Database) | [Bibtex](./iqa_ref.bib#L824-L831)

### No Reference (NR)

- `[Arxiv 2024]` [Q-Mamba: On First Exploration of Vision Mamba for Image Quality Assessment](https://arxiv.org/abs/2406.09546), Guan et al. [Bibtex](./iqa_ref.bib#L980-L986)

- `[Arxiv 2024]` [Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Compare](https://arxiv.org/abs/2405.19298), Zhe et al. [Bibtex](./iqa_ref.bib#L964-L971)

- `[Arxiv 2024]` [Quality-aware Image-Text Alignment for Real-World Image Quality Assessment](https://arxiv.org/abs/2403.11176), Agnolucci et al. [Github](https://github.com/miccunifi/QualiCLIP) | [Bibtex](./iqa_ref.bib#L949-L954)

- `[CVPR 2024]` [Blind Image Quality Assessment Based On Geometric Order Learning](https://openaccess.thecvf.com/content/CVPR2024/papers/Shin_Blind_Image_Quality_Assessment_Based_on_Geometric_Order_Learning_CVPR_2024_paper.pdf), Shin et al. [Github](https://github.com/nhshin-mcl/QCN?tab=readme-ov-file) | [Bibtex](./iqa_ref.bib#L1074-L1079)

- `[CVPR 2024]` [Bridging the Synthetic-to-Authentic Gap: Distortion-Guided Unsupervised Domain Adaptation for Blind Image Quality Assessment](https://arxiv.org/abs/2405.04167), Li et al. [Bibtex](./iqa_ref.bib#L956-L961)

- `[CVPR 2024]` [Boosting Image Quality Assessment through Efficient Transformer Adaptation with Local Feature Enhancement

](https://arxiv.org/abs/2308.12001), Xu et al. | [Bibtex](./iqa_ref.bib#L915-L920)

- `[WACV2024]` [ARNIQA: Learning Distortion Manifold for Image Quality Assessment](https://arxiv.org/abs/2310.14918), Agnolucci et al. [Github](https://github.com/miccunifi/ARNIQA) | [Bibtex](./iqa_ref.bib#L858-L865)

- `[TIP2023]` [TOPIQ: A Top-down Approach from Semantics to Distortions for Image Quality Assessment](https://arxiv.org/abs/2308.03060), Chen et al. [Github](https://github.com/chaofengc/IQA-PyTorch) | [Bibtex](./iqa_ref.bib#L892-L897)

- `[ICCV2023]` [Test Time Adaptation for Blind Image Quality Assessment](https://arxiv.org/abs/2307.14735), Roy et al. [Github](https://github.com/Shankhanil006/TTA-IQA) | [Bibtex](./iqa_ref.bib#L798-L804)

- `[CVPR2023]` [Re-IQA: Unsupervised Learning for Image Quality Assessment in the Wild](https://arxiv.org/abs/2304.00451), Saha et al. [Bibtex](./iqa_ref.bib#L791-L796) | [Github](https://github.com/avinabsaha/ReIQA)

- `[CVPR2023]` [Blind Image Quality Assessment via Vision-Language Correspondence: A Multitask Learning Perspective](https://arxiv.org/abs/2303.14968), Zhang et al. [Github](https://github.com/zwx8981/LIQE) | [Bibtex](./iqa_ref.bib#L770-L775)

- `[CVPR2023]` [Quality-aware Pre-trained Models for Blind Image Quality Assessment](https://arxiv.org/abs/2303.00521), Zhao et al. [Bibtex](./iqa_ref.bib#L763-L768)

- `[AAAI2023]` [Exploring CLIP for Assessing the Look and Feel of Images](https://arxiv.org/abs/2207.12396), Wang et al. [Github](https://github.com/IceClear/CLIP-IQA) | [Bibtex](./iqa_ref.bib#L745-L750)

- `[AAAI2023]` [Data-Efficient Image Quality Assessment with Attention-Panel Decoder](https://arxiv.org/abs/2304.04952), Qin et al. [Github](https://github.com/narthchin/DEIQT) | [Bibtex](./iqa_ref.bib#L745-L750)

- `[TPAMI2022]` [Continual Learning for Blind Image Quality Assessment

](https://arxiv.org/abs/2102.09717), Zhang et al. [Github](https://github.com/zwx8981/BIQA_CL) | [Bibtex](./iqa_ref.bib#L738-L743)

- `[TIP2022]` [No-Reference Image Quality Assessment by Hallucinating Pristine Features](https://arxiv.org/abs/2108.04165), Chen et al. [Github](https://github.com/Baoliang93/FPR) | [Bibtex](./iqa_ref.bib#L848-L856)

- `[TIP2022]` [VCRNet: Visual Compensation Restoration Network for No-Reference Image Quality Assessment](https://ieeexplore.ieee.org/document/9694502), Pan et al. [Github](https://github.com/NUIST-Videocoding/VCRNet) | [Bibtex](./iqa_ref.bib#L752-L761)

- `[TMM2022]` [GraphIQA: Learning Distortion Graph Representations for Blind Image Quality Assessment](https://arxiv.org/abs/2103.07666), Sun et al. [Github](https://github.com/geekyutao/GraphIQA) | [Bibtex](./iqa_ref.bib#L701-L707)

- `[CVPR2021]` [Troubleshooting Blind Image Quality Models in the Wild](https://arxiv.org/abs/2105.06747), Wang et al. [Github](https://github.com/wangzhihua520/troubleshooting_BIQA) | [Bibtex](./iqa_ref.bib#L716-L722)

| Paper Link | Method | Type | Published | Code | Keywords | 

| ----------- | ---------- | ------------| ---------- | ------ | ------ |

| [arXiv](https://arxiv.org/abs/2204.08958) | MANIQA | NR | CVPRW2022 | [Official](https://github.com/IIGROUP/MANIQA) | Transformer, multi-dimension attention, dual branch |

| [arXiv](https://arxiv.org/abs/2108.06858) | TReS | NR | WACV2022 | [Official](https://github.com/isalirezag/TReS) | Transformer, relative ranking, self-consistency |

| [pdf](https://www.bmvc2021-virtualconference.com/assets/papers/0868.pdf) | KonIQ++ | NR | BMVC2021 | [Official](https://github.com/SSL92/koniqplusplus) | Multi-task with distortion prediction 

| [arXiv](https://arxiv.org/abs/2108.05997) | MUSIQ | NR | ICCV2021 | [Official](https://github.com/google-research/google-research/tree/master/musiq) / [Pytorch](https://github.com/anse3832/MUSIQ) | Multi-scale, transformer, Aspect Ratio Preserved (ARP) resizing

| [arXiv](https://arxiv.org/abs/2108.07948) | CKDN | NR | ICCV2021 | [Official](https://github.com/researchmm/CKDN) | Degraded reference, Conditional knowledge distillation (related to HIQA)

| [pdf](https://openaccess.thecvf.com/content_CVPR_2020/papers/Su_Blindly_Assess_Image_Quality_in_the_Wild_Guided_by_a_CVPR_2020_paper.pdf) | HyperIQA | NR | CVPR2020 | [Official](https://github.com/SSL92/hyperIQA) | Content-aware hyper network 

| [arXiv](https://arxiv.org/abs/2004.05508) | Meta-IQA | NR | CVPR2020 | [Official](https://github.com/zhuhancheng/MetaIQA) | Meta-learning 

| [arXiv](https://arxiv.org/abs/2003.08932) | GIQA | NR | ECCV2020 | [Official](https://github.com/cientgu/GIQA) | Generated image 

| [arXiv](https://arxiv.org/abs/1809.07517) | PI | NR | 2018 PIRM Challenge | [Project](https://github.com/roimehrez/PIRM2018) | 1/2 * (NIQE + (10 - NRQM)). 

| [arXiv](https://arxiv.org/abs/1804.01681) | HIQA | NR | CVPR2018 | [Project](https://kwanyeelin.github.io/projects/HIQA/HIQA.html) | Hallucinated reference 

| [arXiv](https://arxiv.org/pdf/1805.08493v1.pdf) | BPSQM | NR | CVPR2018 | []() | Pixel-wise quality map 

| [arXiv](https://arxiv.org/abs/1707.08347) | RankIQA | NR | ICCV2017 | [Github](https://github.com/xialeiliu/RankIQA) | Pretrain on synthetically ranked data 

| [pdf](https://openaccess.thecvf.com/content_cvpr_2014/papers/Kang_Convolutional_Neural_Networks_2014_CVPR_paper.pdf) | CNNIQA | NR | CVPR2014 | [PyTorch](https://github.com/lidq92/CNNIQA) | First CNN-based NR-IQA 

| []() | | | | []() | 

| [arXiv](https://arxiv.org/abs/2005.13983) | UNIQUE | NR | TIP2021 | [Github](https://github.com/zwx8981/UNIQUE) | Combine synthetic and authentic image pairs 

| [arXiv](https://arxiv.org/pdf/1907.02665.pdf) | DBCNN | NR | TCSVT2020 | [Official](https://github.com/zwx8981/DBCNN-PyTorch) | Two branches for synthetic and authentic distortions 

| [pdf](http://www.jdl.link/doc/2011/20191226_08489929.pdf) | SFA | NR | TMM2019 | [Official](https://github.com/lidq92/SFA) | Aggregate ResNet50 features of multiple cropped patches

| [pdf](https://drive.google.com/file/d/1tMjcllKP8SzTn-dWVmogxaCLpzL1L7nO/view)/[arXiv](https://arxiv.org/abs/1708.08190) | PQR | NR/Aesthetic | TIP2019 | [Official1](https://github.com/HuiZeng/Unified_IAA)/[Official2](https://github.com/HuiZeng/BIQA_Toolbox) | Unify different type of aesthetic labels 

| [arXiv](https://arxiv.org/abs/1612.01697) | WaDIQaM (deepIQA) | NR/FR | TIP2018 | [PyTorch](https://github.com/lidq92/WaDIQaM) | Weighted average of patch qualities, shared FR/NR models 

| [pdf](https://ieeexplore.ieee.org/ielx7/83/8347140/08352823.pdf) | NIMA | NR | TIP2018 | [PyTorch](https://github.com/kentsyx/Neural-IMage-Assessment)/[Tensorflow](https://github.com/idealo/image-quality-assessment) | Squared EMD loss 

| [pdf](https://ece.uwaterloo.ca/~z70wang/publications/TIP_E2E_BIQA.pdf) | MEON | NR | TIP2017 | | Multi-task: distortion learning and quality prediction

| [arXiv](https://arxiv.org/abs/1904.06505) | dipIQ | NR | TIP2017 | [download](https://ece.uwaterloo.ca/~k29ma/codes/dipIQ.rar) | Similar to RankIQA

| []() | | | | []() | 

| [arXiv](https://arxiv.org/abs/1612.05890) | NRQM (Ma) | NR | CVIU2017 | [Project](https://sites.google.com/site/chaoma99/sr-metric) | Traditional, Super resolution 

| [arXiv](https://arxiv.org/abs/1609.04757) | FRIQUEE | NR | JoV2017 | [Official](https://github.com/utlive/FRIQUEE) | Authentically Distorted, Bag of Features

| [IEEE](https://ieeexplore.ieee.org/document/7501619) | HOSA | NR | TIP2016 | [Matlab download](https://ieeexplore.ieee.org/document/7501619) | Traditional 

| [pdf](https://live.ece.utexas.edu/publications/2015/zhang2015feature.pdf) | ILNIQE | NR | TIP2015 | [Official](http://www4.comp.polyu.edu.hk/~cslzhang/IQA/ILNIQE/ILNIQE.htm) | Traditional 

| [pdf](https://live.ece.utexas.edu/publications/2012/TIP%20BRISQUE.pdf) | BRISQUE | NR | TIP2012 | [Official](https://github.com/utlive/BRISQUE) | Traditional 

| [pdf](https://live.ece.utexas.edu/publications/2012/saad_2012_tip.pdf) | BLIINDS-II | NR | TIP2012 | [Official](https://github.com/utlive/BLIINDS2) | 

| [pdf](https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.359.7510&rep=rep1&type=pdf) | CORNIA | NR | CVPR2012 | [Matlab download](https://github.com/HuiZeng/BIQA_Toolbox) | Codebook Representation 

| [pdf](https://live.ece.utexas.edu/publications/2013/mittal2013.pdf) | NIQE | NR | SPL2012 | [Official](https://github.com/utlive/niqe) | Traditional 

| [pdf](https://www.imaging.utk.edu/research/wcho/references/2011%20TIP%20BLINDS2.pdf) | DIIVINE | NR | TIP2011 | [Official](https://github.com/utlive/DIIVINE) | 

### Full Reference (FR)

- `[ECCV2022]` [Shift-tolerant Perceptual Similarity Metric](https://arxiv.org/abs/2211.052152207.13686), Ghildyal et al. [Github](https://github.com/abhijay9/ShiftTolerant-LPIPS) | [Bibtex](./iqa_ref.bib#L731-L736)

- `[BMVC2022]` [Content-Diverse Comparisons improve IQA](https://arxiv.org/abs/2211.05215), Thong et al. [Bibtex](./iqa_ref.bib#L724-L729)

- `[ACM MM2022]` [Quality Assessment of Image Super-Resolution: Balancing Deterministic and Statistical Fidelity](https://arxiv.org/abs/2207.08689), Zhou et al. [Github](https://github.com/weizhou-geek/SRIF) | [Bibtex](./iqa_ref.bib#L709-L714)

| Paper Link | Method | Type | Published | Code | Keywords | 

| ----------- | ---------- | ------------| ---------- | ------ | ------ |

| [arXiv](https://arxiv.org/abs/2204.10485) | AHIQ | FR | CVPR2022 NTIRE workshop | [Official](https://github.com/IIGROUP/AHIQ) | Attention, Transformer

| [arXiv](https://arxiv.org/abs/2204.08763) | JSPL | FR | CVPR2022 | [Official](https://github.com/happycaoyue/JSPL) | semi-supervised and positive-unlabeled (PU) learning |

| [arXiv](https://arxiv.org/abs/2202.13123) | CVRKD | NAR | AAAI2022 | [Official](https://github.com/guanghaoyin/CVRKD-IQA) | Non-Aligned content reference, knowledge distillation 

| [arXiv](https://arxiv.org/abs/2104.14730) | IQT | FR | CVPRW2021 | [PyTorch](https://github.com/anse3832/IQT) | Transformer 

| [arXiv](https://arxiv.org/abs/2110.08521) | A-DISTS | FR | ACMM2021 | [Official](https://github.com/dingkeyan93/A-DISTS) | 

| [arXiv](https://arxiv.org/abs/2004.07728) | DISTS | FR | TPAMI2021 | [Official](https://github.com/dingkeyan93/DISTS) | 

| [arXiv](https://arxiv.org/abs/1801.03924) | LPIPS | FR | CVPR2018 | [Project](https://richzhang.github.io/PerceptualSimilarity/) | Perceptual similarity, Pairwise Preference 

| [arXiv](https://arxiv.org/abs/1806.02067) | PieAPP | FR | CVPR2018 | [Project](http://civc.ucsb.edu/graphics/Papers/CVPR2018_PieAPP/) | Perceptual similarity, Pairwise Preference 

| [arXiv](https://arxiv.org/abs/1612.01697) | WaDIQaM | NR/FR | TIP2018 | [Official](https://github.com/lidq92/WaDIQaM) | 

| [arXiv](https://arxiv.org/abs/1902.05316) | JND-SalCAR | FR| TCSVT2020 | []() | JND (Just-Noticeable-Difference) 

| []() | | | | []() | 

| [pdf](https://nottingham-repository.worktribe.com/preview/1589753/Visual%20IEEE-TIP-2019.pdf) | QADS | FR | TIP2019 | [Project](http://www.vista.ac.cn/super-resolution/) | Super-resolution 

| [pdf](https://sse.tongji.edu.cn/linzhang/IQA/FSIM/Files/Fsim%20a%20feature%20similarity%20index%20for%20image%20quality%20assessment.pdf) | FSIM | FR | TIP2011 | [Project](https://sse.tongji.edu.cn/linzhang/IQA/FSIM/FSIM.htm) | Traditional 

| [pdf](https://live.ece.utexas.edu/publications/2004/hrs_ieeetip_2004_imginfo.pdf) | VIF/IFC | FR | TIP2006 | [Project](https://live.ece.utexas.edu/research/Quality/VIF.htm) | Traditional 

| [pdf](https://ece.uwaterloo.ca/~z70wang/publications/msssim.pdf) | MS-SSIM | FR | | [Project](https://ece.uwaterloo.ca/~z70wang/research/ssim/) | Traditional 

| [pdf](https://ece.uwaterloo.ca/~z70wang/publications/ssim.pdf) | SSIM | FR | TIP2004 | [Project](https://ece.uwaterloo.ca/~z70wang/research/ssim/) | Traditional 

| []() | PSNR | FR | | []() | Traditional 

### Image Aesthetic Assessment 

- `[ACMMM 2024]` [AesExpert: Towards Multi-modality Foundation Model for Image Aesthetics Perception](https://arxiv.org/abs/2404.09624), Huang et al. [Project](https://yipoh.github.io/aes-expert/) | [Github](https://github.com/yipoh/AesExpert) | [Bibtex](./iqa_ref.bib#L1037-L1042)

- `[Arxiv 2024]` [AesBench: An Expert Benchmark for Multimodal Large Language Models on Image Aesthetics Perception](https://arxiv.org/abs/2401.08276), Huang et al. [Github](https://github.com/yipoh/AesBench) | [Bibtex](./iqa_ref.bib#L899-L904)

- `[ECCV2024]` [Scaling Up Personalized Aesthetic Assessment via Task Vector Arithmetic](https://arxiv.org/abs/2407.07176), Yun et al. [Bibtex](./iqa_ref.bib#L988-L993) | [Project](https://yeolj00.github.io/personal-projects/personalized-aesthetics/)

- `[CVPR2023]` [VILA: Learning Image Aesthetics from User Comments with Vision-Language Pretraining](https://arxiv.org/abs/2303.14302), Ke et al. [Bibtex](./iqa_ref.bib#L784-L789)

- `[CVPR2023]` [Towards Artistic Image Aesthetics Assessment: a Large-scale Dataset and a New Method](https://arxiv.org/abs/2303.14968), Yi et al. [Github](https://github.com/Dreemurr-T/BAID) | [Bibtex](./iqa_ref.bib#L777-L782)

### Others 

#### Color IQA

- `[ECCV 2024]` [Multiscale Sliced Wasserstein Distances as Perceptual Color Difference Measures](https://arxiv.org/abs/2407.10181), He et al. [Github](https://github.com/real-hjq/MS-SWD) | [Bibtex](./iqa_ref.bib#L1051-L1058)

- `[CVPR 2023]` [Learning a Deep Color Difference Metric for Photographic Images](https://arxiv.org/abs/2303.14964), Chen et al. [Github](https://github.com/haoychen3/CD-Flow) | [Bibtex](./iqa_ref.bib#L1060-L1065)

#### Face IQA

- `[CVPR2024]` [DSL-FIQA: Assessing Facial Image Quality via Dual-Set Degradation Learning and Landmark-Guided Transformer](https://arxiv.org/abs/2406.09622), Chen et al. [Bibtex](./iqa_ref.bib#L973-L978) | [Project](https://dsl-fiqa.github.io/)

#### 360° Image (Omnidirectional Image) IQA

- `[NeurIPS 2023]` [Assessor360: Multi-sequence Network for Blind Omnidirectional Image Quality Assessment

](https://arxiv.org/abs/2305.10983), Wu et al. [Bibtex](./iqa_ref.bib#L995-L1000) | [Github](https://github.com/TianheWu/Assessor360) 

#### Egocentric Spatial Images (Apple Vision Pro)

- `[Arxiv 2024]` [ESIQA: Perceptual Quality Assessment of Vision-Pro-based

Egocentric Spatial Images](https://arxiv.org/abs/2407.21363), Zhu et al. [Bibtex](./iqa_ref.bib#L1009-L1016) | [Github](https://github.com/IntMeGroup/ESIQA)

#### Adversarial Attack on IQA

- `[Arxiv 2024]` [Guardians of Image Quality: Benchmarking Defenses Against Adversarial Attacks on Image Quality Metrics](https://arxiv.org/abs/2408.01541), Gushchin et al. [Bibtex](./iqa_ref.bib#L1027-L1035) | [Github](https://github.com/msu-video-group/adversarial-defenses-for-iqa) | [Project](https://videoprocessing.ai/benchmarks/iqa-defenses.html)

- `[NeurIPS 2022]` [Perceptual Attacks of No-Reference Image Quality Models with Human-in-the-Loop](https://arxiv.org/abs/2210.00933), Zhang et al. [Bibtex](./iqa_ref.bib#L1018-L1025) | [Github](https://github.com/zwx8981/PerceptualAttack_BIQA)

#### IQA Losses

| Title | Method | Published | Code | Keywords | 

| ----------- | ------------| ---------- | ------ | ------ |

| [arXiv](https://arxiv.org/abs/2008.03889) | NiNLoss | ACMM2020 | [Official](https://github.com/lidq92/LinearityIQA) | Norm-in-Norm Loss

## Datasets 

### IQA datasets

| Paper Link | Dataset Name | Type | Published | Website | Images | Annotations

| ----------- | ---------- | ------------| ---------- | ------ | ------ |------ |

| [arXiv](https://arxiv.org/abs/2406.17472) | UHD-IQA | NR | ECCVW2024 | [Project](https://database.mmsp-kn.de/uhd-iqa-benchmark-database.html) | 6k (~3840x2160) | 20 ratings per image |

| [arXiv](https://arxiv.org/abs/1912.10088) | PaQ-2-PiQ | NR | CVPR2020 | [Official github](https://github.com/baidut/PaQ-2-PiQ) | 40k, 120k patches | 4M |

| [CVF](https://openaccess.thecvf.com/content_CVPR_2020/html/Fang_Perceptual_Quality_Assessment_of_Smartphone_Photography_CVPR_2020_paper.html)  | SPAQ | NR | CVPR2020 | [Offical github](https://github.com/h4nwei/SPAQ) | 11k (smartphone) | |  

| [arXiv](https://arxiv.org/abs/1910.06180) | KonIQ-10k | NR | TIP2020 | [Project](http://database.mmsp-kn.de/koniq-10k-database.html) | 10k from [YFCC100M](http://projects.dfki.uni-kl.de/yfcc100m/) | 1.2M | 

| [arXiv](https://arxiv.org/abs/1606.01621v2) | AADB | NR/Aesthentic | ECCV2016 | [Official github](https://github.com/aimerykong/deepImageAestheticsAnalysis) | 10k images (8500/500/1000), 11 attributes 

| [arXiv](https://arxiv.org/abs/1511.02919) | CLIVE | NR | TIP2016 | [Project](https://live.ece.utexas.edu/research/ChallengeDB/index.html) | 1200 | 350k |

| [pdf](http://refbase.cvc.uab.es/files/MMP2012a.pdf) | AVA | NR / Aesthentic | CVPR2012 | [Github](https://github.com/mtobeiyf/ava_downloader)/[Project](http://www.lucamarchesotti.com/) | 250k (60 categories) | | 

| [arXiv](https://arxiv.org/abs/2007.12142) | PIPAL | FR | ECCV2020 | [Project](https://www.jasongt.com/projectpages/pipal.html) | 250 | 1.13M |

| [arXiv](https://arxiv.org/abs/2001.08113) | KADIS-700k | FR | arXiv | [Project](http://database.mmsp-kn.de/kadid-10k-database.html) | 140k pristine / 700k distorted | 30 ratings (DCRs) per image. | 

| [IEEE](https://ieeexplore.ieee.org/document/8743252) | KADID-10k | FR | QoMEX2019 | [Project](http://database.mmsp-kn.de/kadid-10k-database.html) | 81 | 10k distortions |

| [pdf](https://ece.uwaterloo.ca/~k29ma/papers/17_TIP_EXPLORATION.pdf) | Waterloo-Exp | FR | TIP2017 | [Project](https://ece.uwaterloo.ca/~k29ma/exploration/) | 4744 | 94k distortions | 

| [pdf](https://daneshyari.com/article/preview/533080.pdf) | MDID | FR | PR2017 | --- | 20 | 1600 distortions  |

| [pdf](http://www.ponomarenko.info/papers/euvip_tid2013.pdf) | TID2013 | FR | SP2015 | [Project](http://www.ponomarenko.info/tid2013.htm) | 25 | 3000 distortions |

| [pdf](http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.298.9133&rep=rep1&type=pdf)  | LIVEMD | FR | ACSSC2012 | [Project](https://live.ece.utexas.edu/research/Quality/live_multidistortedimage.html) | 15 pristine images | two successive distortions |

| [pdf](https://www.researchgate.net/profile/Damon-Chandler/publication/220050520_Most_apparent_distortion_Full-reference_image_quality_assessment_and_the_role_of_strategy/links/5629cd1c08ae518e347e1445/Most-apparent-distortion-Full-reference-image-quality-assessment-and-the-role-of-strategy.pdf)  | CSIQ | FR | Journal of Electronic Imaging 2010 | --- | 30 | 866 distortions |

| [pdf](http://www.ponomarenko.info/papers/mre2009tid.pdf) | TID2008 | FR | 2009 | [Project](http://www.ponomarenko.info/tid2008.htm) | 25 | 1700 distortions |

| [pdf](https://live.ece.utexas.edu/publications/2006/hrs-transIP-06.pdf) | LIVE IQA | FR | TIP2006 | [Project](https://live.ece.utexas.edu/research/Quality/subjective.htm) | 29 images, 780 synthetic distortions 

| [link](http://hal.univ-nantes.fr/hal-00580755/) | IVC | FR | 2005 | --- | 10 | 185 distortions |

### Perceptual similarity datasets

| Paper Title | Dataset Name | Type | Published | Website | Images | Annotations |  

| ----------- | ---------- | ------------| ---------- | ------ | ------ | ----- |

| [arXiv](https://arxiv.org/abs/1801.03924) | BAPPS(LPIPS) | FR | CVPR2018 | [Project](https://richzhang.github.io/PerceptualSimilarity/) | 187.7k  | 484k 

| [arXiv](https://arxiv.org/abs/1806.02067) | PieAPP | FR | CVPR2018 | [Project](http://civc.ucsb.edu/graphics/Papers/CVPR2018_PieAPP/) | 200 images | 2.3M