{"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/cnn-features-off-the-shelf-an-astounding","title":"CNN Features off-the-shelf: an Astounding Baseline for Recognition","arxiv_id":"1403.6382","date":"2014-03-23","proceeding":null,"authors":["Ali Sharif Razavian","Hossein Azizpour","Josephine Sullivan","Stefan Carlsson"],"abstract":"Recent results indicate that the generic descriptors extracted from the\nconvolutional neural networks are very powerful. This paper adds to the\nmounting evidence that this is indeed the case. We report on a series of\nexperiments conducted for different recognition tasks using the publicly\navailable code and model of the \\overfeat network which was trained to perform\nobject classification on ILSVRC13. We use features extracted from the \\overfeat\nnetwork as a generic image representation to tackle the diverse range of\nrecognition tasks of object image classification, scene recognition, fine\ngrained recognition, attribute detection and image retrieval applied to a\ndiverse set of datasets. We selected these tasks and datasets as they gradually\nmove further away from the original task and data the \\overfeat network was\ntrained to solve. Astonishingly, we report consistent superior results compared\nto the highly tuned state-of-the-art systems in all the visual classification\ntasks on various datasets. For instance retrieval it consistently outperforms\nlow memory footprint methods except for sculptures dataset. The results are\nachieved using a linear SVM classifier (or $L2$ distance in case of retrieval)\napplied to a feature representation of size 4096 extracted from a layer in the\nnet. The representations are further modified using simple augmentation\ntechniques e.g. jittering. The results strongly suggest that features obtained\nfrom deep learning with convolutional nets should be the primary candidate in\nmost visual recognition tasks.","url_abs":"http://arxiv.org/abs/1403.6382v3","url_pdf":"http://arxiv.org/pdf/1403.6382v3.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":"cnn-features-off-the-shelf-an-astounding","repo_url":"https://github.com/Aniket7/Transfer-Learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"cnn-features-off-the-shelf-an-astounding","repo_url":"https://github.com/baldassarreFe/deep-koalarization","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"cnn-features-off-the-shelf-an-astounding","repo_url":"https://github.com/jimgoo/caffe-oxford102","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"cnn-features-off-the-shelf-an-astounding","repo_url":"https://github.com/joshuaczhao/CNN-Sentence-Classifier-Reproduction","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"attribute","task_name":"Attribute"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-retrieval","task_name":"Image Retrieval"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"scene-recognition","task_name":"Scene Recognition"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"svm","method_name":"SVM"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1403.6382","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}