{"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/evaluation-of-deep-convolutional-nets-for","title":"Evaluation of Deep Convolutional Nets for Document Image Classification and Retrieval","arxiv_id":"1502.07058","date":"2015-02-25","proceeding":null,"authors":["Adam W. Harley","Alex Ufkes","Konstantinos G. Derpanis"],"abstract":"This paper presents a new state-of-the-art for document image classification\nand retrieval, using features learned by deep convolutional neural networks\n(CNNs). In object and scene analysis, deep neural nets are capable of learning\na hierarchical chain of abstraction from pixel inputs to concise and\ndescriptive representations. The current work explores this capacity in the\nrealm of document analysis, and confirms that this representation strategy is\nsuperior to a variety of popular hand-crafted alternatives. Experiments also\nshow that (i) features extracted from CNNs are robust to compression, (ii) CNNs\ntrained on non-document images transfer well to document analysis tasks, and\n(iii) enforcing region-specific feature-learning is unnecessary given\nsufficient training data. This work also makes available a new labelled subset\nof the IIT-CDIP collection, containing 400,000 document images across 16\ncategories, useful for training new CNNs for document analysis.","url_abs":"http://arxiv.org/abs/1502.07058v1","url_pdf":"http://arxiv.org/pdf/1502.07058v1.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":"descriptive","task_name":"Descriptive"},{"task_slug":"document-image-classification","task_name":"Document Image Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"document-image-classification","task_name":"document-image-classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[{"slug":"rvl-cdip","name":"RVL-CDIP","full_name":"RVL-CDIP"}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1502.07058","atlas_url":"https://app.syntology.ai/?focus=1502.07058","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}