{"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/recurrent-computations-for-visual-pattern","title":"Recurrent computations for visual pattern completion","arxiv_id":"1706.02240","date":"2017-06-07","proceeding":null,"authors":["Hanlin Tang","Martin Schrimpf","Bill Lotter","Charlotte Moerman","Ana Paredes","Josue Ortega Caro","Walter Hardesty","David Cox","Gabriel Kreiman"],"abstract":"Making inferences from partial information constitutes a critical aspect of\ncognition. During visual perception, pattern completion enables recognition of\npoorly visible or occluded objects. We combined psychophysics, physiology and\ncomputational models to test the hypothesis that pattern completion is\nimplemented by recurrent computations and present three pieces of evidence that\nare consistent with this hypothesis. First, subjects robustly recognized\nobjects even when rendered <15% visible, but recognition was largely impaired\nwhen processing was interrupted by backward masking. Second, invasive\nphysiological responses along the human ventral cortex exhibited visually\nselective responses to partially visible objects that were delayed compared to\nwhole objects, suggesting the need for additional computations. These\nphysiological delays were correlated with the effects of backward masking.\nThird, state-of-the-art feed-forward computational architectures were not\nrobust to partial visibility. However, recognition performance was recovered\nwhen the model was augmented with attractor-based recurrent connectivity. These\nresults provide a strong argument of plausibility for the role of recurrent\ncomputations in making visual inferences from partial information.","url_abs":"http://arxiv.org/abs/1706.02240v2","url_pdf":"http://arxiv.org/pdf/1706.02240v2.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":"recurrent-computations-for-visual-pattern","repo_url":"https://github.com/kreimanlab/occlusion-classification","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1706.02240","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}