{"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/grad-cam-why-did-you-say-that","title":"Grad-CAM: Why did you say that?","arxiv_id":"1611.07450","date":"2016-11-22","proceeding":null,"authors":["Ramprasaath R. Selvaraju","Abhishek Das","Ramakrishna Vedantam","Michael Cogswell","Devi Parikh","Dhruv Batra"],"abstract":"We propose a technique for making Convolutional Neural Network (CNN)-based\nmodels more transparent by visualizing input regions that are 'important' for\npredictions -- or visual explanations. Our approach, called Gradient-weighted\nClass Activation Mapping (Grad-CAM), uses class-specific gradient information\nto localize important regions. These localizations are combined with existing\npixel-space visualizations to create a novel high-resolution and\nclass-discriminative visualization called Guided Grad-CAM. These methods help\nbetter understand CNN-based models, including image captioning and visual\nquestion answering (VQA) models. We evaluate our visual explanations by\nmeasuring their ability to discriminate between classes, to inspire trust in\nhumans, and their correlation with occlusion maps. Grad-CAM provides a new way\nto understand CNN-based models.\n  We have released code, an online demo hosted on CloudCV, and a full version\nof this extended abstract.","url_abs":"http://arxiv.org/abs/1611.07450v2","url_pdf":"http://arxiv.org/pdf/1611.07450v2.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":"grad-cam-why-did-you-say-that","repo_url":"https://github.com/ramprs/grad-cam","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"torch","reach":{"status":"ok"}},{"paper_slug":"grad-cam-why-did-you-say-that","repo_url":"https://github.com/jacobgil/pytorch-grad-cam","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-captioning","task_name":"Image Captioning"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1611.07450","atlas_url":"https://app.syntology.ai/?focus=1611.07450","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}