{"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/training-convolutional-neural-networks-with","title":"Training convolutional neural networks with megapixel images","arxiv_id":"1804.05712","date":"2018-04-16","proceeding":null,"authors":["Hans Pinckaers","Geert Litjens"],"abstract":"To train deep convolutional neural networks, the input data and the\nintermediate activations need to be kept in memory to calculate the gradient\ndescent step. Given the limited memory available in the current generation\naccelerator cards, this limits the maximum dimensions of the input data. We\ndemonstrate a method to train convolutional neural networks holding only parts\nof the image in memory while giving equivalent results. We quantitatively\ncompare this new way of training convolutional neural networks with\nconventional training. In addition, as a proof of concept, we train a\nconvolutional neural network with 64 megapixel images, which requires 97% less\nmemory than the conventional approach.","url_abs":"http://arxiv.org/abs/1804.05712v1","url_pdf":"http://arxiv.org/pdf/1804.05712v1.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":"training-convolutional-neural-networks-with","repo_url":"https://github.com/DIAGNijmegen/StreamingSGD","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1804.05712","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}