{"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/counting-everyday-objects-in-everyday-scenes","title":"Counting Everyday Objects in Everyday Scenes","arxiv_id":"1604.03505","date":"2016-04-12","proceeding":"CVPR 2017 7","authors":["Prithvijit Chattopadhyay","Ramakrishna Vedantam","Ramprasaath R. Selvaraju","Dhruv Batra","Devi Parikh"],"abstract":"We are interested in counting the number of instances of object classes in\nnatural, everyday images. Previous counting approaches tackle the problem in\nrestricted domains such as counting pedestrians in surveillance videos. Counts\ncan also be estimated from outputs of other vision tasks like object detection.\nIn this work, we build dedicated models for counting designed to tackle the\nlarge variance in counts, appearances, and scales of objects found in natural\nscenes. Our approach is inspired by the phenomenon of subitizing - the ability\nof humans to make quick assessments of counts given a perceptual signal, for\nsmall count values. Given a natural scene, we employ a divide and conquer\nstrategy while incorporating context across the scene to adapt the subitizing\nidea to counting. Our approach offers consistent improvements over numerous\nbaseline approaches for counting on the PASCAL VOC 2007 and COCO datasets.\nSubsequently, we study how counting can be used to improve object detection. We\nthen show a proof of concept application of our counting methods to the task of\nVisual Question Answering, by studying the `how many?' questions in the VQA and\nCOCO-QA datasets.","url_abs":"http://arxiv.org/abs/1604.03505v3","url_pdf":"http://arxiv.org/pdf/1604.03505v3.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":"counting-everyday-objects-in-everyday-scenes","repo_url":"https://github.com/prithv1/cvpr2017_counting","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"torch","reach":null}],"tasks":[{"task_slug":"object","task_name":"Object"},{"task_slug":"object-counting","task_name":"Object Counting"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/object-counting-on-coco-count-test","task":"Object Counting","dataset":"COCO count-test","model":"ens","rank_in_archive_order":1,"of":7,"metrics":{"m-reIRMSE":"0.18","m-reIRMSE-nz":"0.81","mRMSE":"0.36","mRMSE-nz":"1.98"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-coco-count-test","task":"Object Counting","dataset":"COCO count-test","model":"Seq-sub-ft-3x3","rank_in_archive_order":2,"of":7,"metrics":{"m-reIRMSE":"0.18","m-reIRMSE-nz":"0.82","mRMSE":"0.35","mRMSE-nz":"1.96"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-coco-count-test","task":"Object Counting","dataset":"COCO count-test","model":"Fast-RCNN","rank_in_archive_order":5,"of":7,"metrics":{"m-reIRMSE":"0.20","m-reIRMSE-nz":"1.13","mRMSE":"0.49","mRMSE-nz":"2.78"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-coco-count-test","task":"Object Counting","dataset":"COCO count-test","model":"glance-ft-2L","rank_in_archive_order":6,"of":7,"metrics":{"m-reIRMSE":"0.23","m-reIRMSE-nz":"0.91","mRMSE":"0.42","mRMSE-nz":"2.25"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-coco-count-test","task":"Object Counting","dataset":"COCO count-test","model":"Aso-sub-ft-3x3","rank_in_archive_order":7,"of":7,"metrics":{"m-reIRMSE":"0.24","m-reIRMSE-nz":"0.87","mRMSE":"0.38","mRMSE-nz":"2.08"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-pascal-voc","task":"Object Counting","dataset":"PASCAL VOC","model":"CEOES","rank_in_archive_order":3,"of":3,"metrics":{"mRMSE":"0.42"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-pascal-voc-2007-count-test","task":"Object Counting","dataset":"Pascal VOC 2007 count-test","model":"ens","rank_in_archive_order":3,"of":8,"metrics":{"m-reIRMSE-nz":"0.65","m-relRMSE":"0.20","mRMSE":"0.42","mRMSE-nz":"1.68"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-pascal-voc-2007-count-test","task":"Object Counting","dataset":"Pascal VOC 2007 count-test","model":"Seq-sub-ft-3x3","rank_in_archive_order":4,"of":8,"metrics":{"m-reIRMSE-nz":"0.68","m-relRMSE":"0.22","mRMSE":"0.43","mRMSE-nz":"1.65"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-pascal-voc-2007-count-test","task":"Object Counting","dataset":"Pascal VOC 2007 count-test","model":"glance-noft-2L","rank_in_archive_order":6,"of":8,"metrics":{"m-reIRMSE-nz":"0.73","m-relRMSE":"0.27","mRMSE":"0.50","mRMSE-nz":"1.83"},"uses_additional_data":false},{"leaderboard":"/sota/object-counting-on-pascal-voc-2007-count-test","task":"Object Counting","dataset":"Pascal VOC 2007 count-test","model":"Fast-RCNN","rank_in_archive_order":7,"of":8,"metrics":{"m-reIRMSE-nz":"0.85","m-relRMSE":"0.26","mRMSE":"0.50","mRMSE-nz":"1.92"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1604.03505","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}