{"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/compressing-the-input-for-cnns-with-the-first","title":"Compressing the Input for CNNs with the First-Order Scattering Transform","arxiv_id":"1809.10200","date":"2018-09-27","proceeding":"ECCV 2018 9","authors":["Edouard Oyallon","Eugene Belilovsky","Sergey Zagoruyko","Michal Valko"],"abstract":"We study the first-order scattering transform as a candidate for reducing the\nsignal processed by a convolutional neural network (CNN). We show theoretical\nand empirical evidence that in the case of natural images and sufficiently\nsmall translation invariance, this transform preserves most of the signal\ninformation needed for classification while substantially reducing the spatial\nresolution and total signal size. We demonstrate that cascading a CNN with this\nrepresentation performs on par with ImageNet classification models, commonly\nused in downstream tasks, such as the ResNet-50. We subsequently apply our\ntrained hybrid ImageNet model as a base model on a detection system, which has\ntypically larger image inputs. On Pascal VOC and COCO detection tasks we\ndemonstrate improvements in the inference speed and training memory consumption\ncompared to models trained directly on the input image.","url_abs":"http://arxiv.org/abs/1809.10200v1","url_pdf":"http://arxiv.org/pdf/1809.10200v1.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":"compressing-the-input-for-cnns-with-the-first","repo_url":"https://github.com/edouardoyallon/pyscatlight","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"translation","task_name":"Translation"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1809.10200","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}