{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/koniq-10k-towards-an-ecologically-valid-and","title":"KonIQ-10k: Towards an ecologically valid and large-scale IQA database","arxiv_id":"1803.08489","date":"2018-03-22","proceeding":null,"authors":["Hanhe Lin","Vlad Hosu","Dietmar Saupe"],"abstract":"The main challenge in applying state-of-the-art deep learning methods to\npredict image quality in-the-wild is the relatively small size of existing\nquality scored datasets. The reason for the lack of larger datasets is the\nmassive resources required in generating diverse and publishable content. We\npresent a new systematic and scalable approach to create large-scale, authentic\nand diverse image datasets for Image Quality Assessment (IQA). We show how we\nbuilt an IQA database, KonIQ-10k, consisting of 10,073 images, on which we\nperformed very large scale crowdsourcing experiments in order to obtain\nreliable quality ratings from 1,467 crowd workers (1.2 million ratings). We\nargue for its ecological validity by analyzing the diversity of the dataset, by\ncomparing it to state-of-the-art IQA databases, and by checking the reliability\nof our user studies.","url_abs":"http://arxiv.org/abs/1803.08489v1","url_pdf":"http://arxiv.org/pdf/1803.08489v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"koniq-10k-towards-an-ecologically-valid-and","repo_url":"https://github.com/subpic/koniq","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"diversity","task_name":"Diversity"},{"task_slug":"image-quality-assessment","task_name":"Image Quality Assessment"},{"task_slug":null,"task_name":"valid"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-quality-assessment-on-koniq-10k","task":"Image Quality Assessment","dataset":"KonIQ-10k","model":"KonCept512","rank_in_archive_order":4,"of":4,"metrics":{"SRCC":"0.921"},"uses_additional_data":false},{"leaderboard":"/sota/image-quality-assessment-on-msu-nr-vqa","task":"Image Quality Assessment","dataset":"MSU NR VQA Database","model":"KonCept512","rank_in_archive_order":9,"of":10,"metrics":{"KLCC":"0.6608","PLCC":"0.8464","SRCC":"0.8360"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1803.08489","atlas_url":"https://app.syntology.ai/?focus=1803.08489","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}