{"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/scanner-efficient-video-analysis-at-scale","title":"Scanner: Efficient Video Analysis at Scale","arxiv_id":"1805.07339","date":"2018-05-18","proceeding":null,"authors":["Alex Poms","Will Crichton","Pat Hanrahan","Kayvon Fatahalian"],"abstract":"A growing number of visual computing applications depend on the analysis of\nlarge video collections. The challenge is that scaling applications to operate\non these datasets requires efficient systems for pixel data access and parallel\nprocessing across large numbers of machines. Few programmers have the\ncapability to operate efficiently at these scales, limiting the field's ability\nto explore new applications that leverage big video data. In response, we have\ncreated Scanner, a system for productive and efficient video analysis at scale.\nScanner organizes video collections as tables in a data store optimized for\nsampling frames from compressed video, and executes pixel processing\ncomputations, expressed as dataflow graphs, on these frames. Scanner schedules\nvideo analysis applications expressed using these abstractions onto\nheterogeneous throughput computing hardware, such as multi-core CPUs, GPUs, and\nmedia processing ASICs, for high-throughput pixel processing. We demonstrate\nthe productivity of Scanner by authoring a variety of video processing\napplications including the synthesis of stereo VR video streams from\nmulti-camera rigs, markerless 3D human pose reconstruction from video, and\ndata-mining big video datasets such as hundreds of feature-length films or over\n70,000 hours of TV news. These applications achieve near-expert performance on\na single machine and scale efficiently to hundreds of machines, enabling\nformerly long-running big video data analysis tasks to be carried out in\nminutes to hours.","url_abs":"http://arxiv.org/abs/1805.07339v1","url_pdf":"http://arxiv.org/pdf/1805.07339v1.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":"scanner-efficient-video-analysis-at-scale","repo_url":"https://github.com/scanner-research/scanner","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}