{"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/aggregation-cross-entropy-for-sequence","title":"Aggregation Cross-Entropy for Sequence Recognition","arxiv_id":"1904.08364","date":"2019-04-17","proceeding":"CVPR 2019 6","authors":["Zecheng Xie","Yaoxiong Huang","Yuanzhi Zhu","Lianwen Jin","Yuliang Liu","Lele Xie"],"abstract":"In this paper, we propose a novel method, aggregation cross-entropy (ACE),\nfor sequence recognition from a brand new perspective. The ACE loss function\nexhibits competitive performance to CTC and the attention mechanism, with much\nquicker implementation (as it involves only four fundamental formulas), faster\ninference\\back-propagation (approximately O(1) in parallel), less storage\nrequirement (no parameter and negligible runtime memory), and convenient\nemployment (by replacing CTC with ACE). Furthermore, the proposed ACE loss\nfunction exhibits two noteworthy properties: (1) it can be directly applied for\n2D prediction by flattening the 2D prediction into 1D prediction as the input\nand (2) it requires only characters and their numbers in the sequence\nannotation for supervision, which allows it to advance beyond sequence\nrecognition, e.g., counting problem. The code is publicly available at\nhttps://github.com/summerlvsong/Aggregation-Cross-Entropy.","url_abs":"http://arxiv.org/abs/1904.08364v2","url_pdf":"http://arxiv.org/pdf/1904.08364v2.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":"aggregation-cross-entropy-for-sequence","repo_url":"https://github.com/summerlvsong/Aggregation-Cross-Entropy","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"aggregation-cross-entropy-for-sequence","repo_url":"https://github.com/myhub/notes","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"prediction","task_name":"Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.08364","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}