{"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modified by Syntology; see the attribution page"},"url":"/paper/historical-document-image-segmentation-with","title":"Historical Document Image Segmentation with LDA-Initialized Deep Neural Networks","arxiv_id":"1710.07363","date":"2017-10-19","proceeding":null,"authors":["Michele Alberti","Mathias Seuret","Vinaychandran Pondenkandath","Rolf Ingold","Marcus Liwicki"],"abstract":"In this paper, we present a novel approach to perform deep neural networks\nlayer-wise weight initialization using Linear Discriminant Analysis (LDA).\nTypically, the weights of a deep neural network are initialized with: random\nvalues, greedy layer-wise pre-training (usually as Deep Belief Network or as\nauto-encoder) or by re-using the layers from another network (transfer\nlearning). Hence, many training epochs are needed before meaningful weights are\nlearned, or a rather similar dataset is required for seeding a fine-tuning of\ntransfer learning. In this paper, we describe how to turn an LDA into either a\nneural layer or a classification layer. We analyze the initialization technique\non historical documents. First, we show that an LDA-based initialization is\nquick and leads to a very stable initialization. Furthermore, for the task of\nlayout analysis at pixel level, we investigate the effectiveness of LDA-based\ninitialization and show that it outperforms state-of-the-art random weight\ninitialization methods.","url_abs":"http://arxiv.org/abs/1710.07363v1","url_pdf":"http://arxiv.org/pdf/1710.07363v1.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":"historical-document-image-segmentation-with","repo_url":"https://github.com/DIVA-DIA/LayoutAnalysisEvaluator","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"image-segmentation","task_name":"Image Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[{"method_slug":"deep-belief-network","method_name":"Deep Belief Network"},{"method_slug":"lda","method_name":"LDA"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}