{"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/learning-photography-aesthetics-with-deep","title":"Learning Photography Aesthetics with Deep CNNs","arxiv_id":"1707.03981","date":"2017-07-13","proceeding":null,"authors":["Gautam Malu","Raju S. Bapi","Bipin Indurkhya"],"abstract":"Automatic photo aesthetic assessment is a challenging artificial intelligence\ntask. Existing computational approaches have focused on modeling a single\naesthetic score or a class (good or bad), however these do not provide any\ndetails on why the photograph is good or bad, or which attributes contribute to\nthe quality of the photograph. To obtain both accuracy and human interpretation\nof the score, we advocate learning the aesthetic attributes along with the\nprediction of the overall score. For this purpose, we propose a novel multitask\ndeep convolution neural network, which jointly learns eight aesthetic\nattributes along with the overall aesthetic score. We report near human\nperformance in the prediction of the overall aesthetic score. To understand the\ninternal representation of these attributes in the learned model, we also\ndevelop the visualization technique using back propagation of gradients. These\nvisualizations highlight the important image regions for the corresponding\nattributes, thus providing insights about model's representation of these\nattributes. We showcase the diversity and complexity associated with different\nattributes through a qualitative analysis of the activation maps.","url_abs":"http://arxiv.org/abs/1707.03981v1","url_pdf":"http://arxiv.org/pdf/1707.03981v1.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":"learning-photography-aesthetics-with-deep","repo_url":"https://github.com/gautamMalu/Aesthetic_attributes_maps","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null},{"paper_slug":"learning-photography-aesthetics-with-deep","repo_url":"https://github.com/kevinlu1211/deep-photo-aesthetics","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"learning-photography-aesthetics-with-deep","repo_url":"https://github.com/rawmarshmellows/deep-photo-aesthetics","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"diversity","task_name":"Diversity"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1707.03981","atlas_url":"https://app.syntology.ai/?focus=1707.03981","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}