{"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/cake-compact-and-accurate-k-dimensional","title":"CAKE: Compact and Accurate K-dimensional representation of Emotion","arxiv_id":"1807.11215","date":"2018-07-30","proceeding":null,"authors":["Corentin Kervadec","Valentin Vielzeuf","Stéphane Pateux","Alexis Lechervy","Frédéric Jurie"],"abstract":"Numerous models describing the human emotional states have been built by the\npsychology community. Alongside, Deep Neural Networks (DNN) are reaching\nexcellent performances and are becoming interesting features extraction tools\nin many computer vision tasks.Inspired by works from the psychology community,\nwe first study the link between the compact two-dimensional representation of\nthe emotion known as arousal-valence, and discrete emotion classes (e.g. anger,\nhappiness, sadness, etc.) used in the computer vision community. It enables to\nassess the benefits -- in terms of discrete emotion inference -- of adding an\nextra dimension to arousal-valence (usually named dominance). Building on these\nobservations, we propose CAKE, a 3-dimensional representation of emotion\nlearned in a multi-domain fashion, achieving accurate emotion recognition on\nseveral public datasets. Moreover, we visualize how emotions boundaries are\norganized inside DNN representations and show that DNNs are implicitly learning\narousal-valence-like descriptions of emotions. Finally, we use the CAKE\nrepresentation to compare the quality of the annotations of different public\ndatasets.","url_abs":"http://arxiv.org/abs/1807.11215v2","url_pdf":"http://arxiv.org/pdf/1807.11215v2.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":[],"tasks":[{"task_slug":"emotion-recognition","task_name":"Emotion Recognition"},{"task_slug":"facial-expression-recognition","task_name":"Facial Expression Recognition (FER)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/facial-expression-recognition-on-affectnet","task":"Facial Expression Recognition (FER)","dataset":"AffectNet","model":"CAKE","rank_in_archive_order":49,"of":50,"metrics":{"Accuracy (7 emotion)":"61.7"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1807.11215","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}