{"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/predicting-aesthetic-score-distribution","title":"Predicting Aesthetic Score Distribution through Cumulative Jensen-Shannon Divergence","arxiv_id":"1708.07089","date":"2017-08-23","proceeding":null,"authors":["Xin Jin","Le Wu","Xiao-Dong Li","Siyu Chen","Siwei Peng","Jingying Chi","Shiming Ge","Chenggen Song","Geng Zhao"],"abstract":"Aesthetic quality prediction is a challenging task in the computer vision\ncommunity because of the complex interplay with semantic contents and\nphotographic technologies. Recent studies on the powerful deep learning based\naesthetic quality assessment usually use a binary high-low label or a numerical\nscore to represent the aesthetic quality. However the scalar representation\ncannot describe well the underlying varieties of the human perception of\naesthetics. In this work, we propose to predict the aesthetic score\ndistribution (i.e., a score distribution vector of the ordinal basic human\nratings) using Deep Convolutional Neural Network (DCNN). Conventional DCNNs\nwhich aim to minimize the difference between the predicted scalar numbers or\nvectors and the ground truth cannot be directly used for the ordinal basic\nrating distribution. Thus, a novel CNN based on the Cumulative distribution\nwith Jensen-Shannon divergence (CJS-CNN) is presented to predict the aesthetic\nscore distribution of human ratings, with a new reliability-sensitive learning\nmethod based on the kurtosis of the score distribution, which eliminates the\nrequirement of the original full data of human ratings (without normalization).\nExperimental results on large scale aesthetic dataset demonstrate the\neffectiveness of our introduced CJS-CNN in this task.","url_abs":"http://arxiv.org/abs/1708.07089v2","url_pdf":"http://arxiv.org/pdf/1708.07089v2.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":"predicting-aesthetic-score-distribution","repo_url":"https://github.com/BestiVictory/CJS-CNN","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"predicting-aesthetic-score-distribution","repo_url":"https://github.com/luke321321/portfolio","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.07089","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}