{"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/between-class-learning-for-image","title":"Between-class Learning for Image Classification","arxiv_id":"1711.10284","date":"2017-11-28","proceeding":"CVPR 2018 6","authors":["Yuji Tokozume","Yoshitaka Ushiku","Tatsuya Harada"],"abstract":"In this paper, we propose a novel learning method for image classification\ncalled Between-Class learning (BC learning). We generate between-class images\nby mixing two images belonging to different classes with a random ratio. We\nthen input the mixed image to the model and train the model to output the\nmixing ratio. BC learning has the ability to impose constraints on the shape of\nthe feature distributions, and thus the generalization ability is improved. BC\nlearning is originally a method developed for sounds, which can be digitally\nmixed. Mixing two image data does not appear to make sense; however, we argue\nthat because convolutional neural networks have an aspect of treating input\ndata as waveforms, what works on sounds must also work on images. First, we\npropose a simple mixing method using internal divisions, which surprisingly\nproves to significantly improve performance. Second, we propose a mixing method\nthat treats the images as waveforms, which leads to a further improvement in\nperformance. As a result, we achieved 19.4% and 2.26% top-1 errors on\nImageNet-1K and CIFAR-10, respectively.","url_abs":"http://arxiv.org/abs/1711.10284v2","url_pdf":"http://arxiv.org/pdf/1711.10284v2.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":"between-class-learning-for-image","repo_url":"https://github.com/mil-tokyo/bc_learning_image","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"torch","reach":{"status":"unanswered"}},{"paper_slug":"between-class-learning-for-image","repo_url":"https://github.com/mil-tokyo/bc_learning_sound","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"between-class-learning-for-image","repo_url":"https://github.com/rishabh135/foodx","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1711.10284","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}