{"id":"https://openalex.org/W3203774546","doi":"https://doi.org/10.48550/arxiv.2109.13227","title":"TSM: Temporal Shift Module for Efficient and Scalable Video Understanding on Edge Device","display_name":"TSM: Temporal Shift Module for Efficient and Scalable Video Understanding on Edge Device","publication_year":2021,"publication_date":"2021-09-27","ids":{"openalex":"https://openalex.org/W3203774546","doi":"https://doi.org/10.48550/arxiv.2109.13227","mag":"3203774546"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2109.13227","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2109.13227","pdf_url":"https://arxiv.org/pdf/2109.13227","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2109.13227","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5065001783","display_name":"Ji Lin","orcid":"https://orcid.org/0000-0001-6053-4344","observed_orcids":["https://orcid.org/0000-0001-6053-4344"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Ji","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040877128","display_name":"Chuang Gan","orcid":"https://orcid.org/0000-0003-4031-5886","observed_orcids":["https://orcid.org/0000-0003-4031-5886"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gan, Chuang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100707345","display_name":"Kuan Wang","orcid":"https://orcid.org/0000-0002-4420-6620","observed_orcids":["https://orcid.org/0000-0002-4420-6620"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Kuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5070926896","display_name":"Song Han","orcid":"https://orcid.org/0000-0002-4186-7618","observed_orcids":["https://orcid.org/0000-0002-4186-7618"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Song","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10812","display_name":"Human Pose and Action Recognition","score":0.9926999807357788,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10812","display_name":"Human Pose and Action Recognition","score":0.9926999807357788,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.005400000140070915,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11398","display_name":"Hand Gesture Recognition Systems","score":0.0005000000237487257,"subfield":{"id":"https://openalex.org/subfields/1709","display_name":"Human-Computer Interaction"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/temporal-shift-module","display_name":"temporal shift module","score":0.9390000104904175},{"id":"https://openalex.org/keywords/video-understanding","display_name":"video understanding","score":0.7260000109672546},{"id":"https://openalex.org/keywords/human-action-recognition","display_name":"human action recognition","score":0.6840000152587891},{"id":"https://openalex.org/keywords/temporal-modeling","display_name":"temporal modeling","score":0.7720000147819519},{"id":"https://openalex.org/keywords/2d-convolutional-neural-network","display_name":"2D convolutional neural network","score":0.7080000042915344},{"id":"https://openalex.org/keywords/edge-devices","display_name":"edge devices","score":0.7239999771118164},{"id":"https://openalex.org/keywords/nvidia-jetson-nano","display_name":"Jetson Nano","score":0.7419999837875366},{"id":"https://openalex.org/keywords/scalability","display_name":"scalability","score":0.5740000009536743},{"id":"https://openalex.org/keywords/kinetics","display_name":"kinetics","score":0.9649999737739563}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7972643971443176},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.7066973447799683},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6690257787704468},{"id":"https://openalex.org/C3261483","wikidata":"https://www.wikidata.org/wiki/Q119565","display_name":"Frame rate","level":2,"score":0.5429766178131104},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5001192092895508},{"id":"https://openalex.org/C162307627","wikidata":"https://www.wikidata.org/wiki/Q204833","display_name":"Enhanced Data Rates for GSM Evolution","level":2,"score":0.49154552817344666},{"id":"https://openalex.org/C33676613","wikidata":"https://www.wikidata.org/wiki/Q13415176","display_name":"Dimension (graph theory)","level":2,"score":0.4491446614265442},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.44099748134613037},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4381495714187622},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.42844706773757935},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.182331383228302},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.0},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.0},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2109.13227","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2109.13227","pdf_url":"https://arxiv.org/pdf/2109.13227","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2109.13227","doi":"https://doi.org/10.48550/arxiv.2109.13227","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2109.13227","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2109.13227","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2109.13227","pdf_url":"https://arxiv.org/pdf/2109.13227","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2389214306","https://openalex.org/W4235240664","https://openalex.org/W2965083567","https://openalex.org/W1838576100","https://openalex.org/W2095886385","https://openalex.org/W2889616422","https://openalex.org/W2089704382","https://openalex.org/W1983399550","https://openalex.org/W97075385","https://openalex.org/W2100349471"],"abstract_inverted_index":{"The":[0,59,193],"explosive":[1],"growth":[2],"in":[3,160],"video":[4,7,136],"streaming":[5],"requires":[6],"understanding":[8],"at":[9,93,198],"high":[10,54,57,107,122,127],"accuracy":[11],"and":[12,45,56,96,132,141,180],"low":[13],"computation":[14,95],"cost.":[15],"Conventional":[16],"2D":[17,87,170],"CNNs":[18,88],"are":[19,35],"computationally":[20,36],"cheap":[21],"but":[22,34],"cannot":[23,172],"capture":[24],"temporal":[25,73,91],"relationships;":[26],"3D":[27,151],"CNN-based":[28],"methods":[29],"can":[30,83],"achieve":[31,90],"good":[32],"performance":[33],"intensive.":[37],"In":[38],"this":[39],"paper,":[40],"we":[41,174],"propose":[42],"a":[43,126],"generic":[44],"effective":[46],"Temporal":[47],"Shift":[48],"Module":[49],"(TSM)":[50],"that":[51,182],"enjoys":[52],"both":[53],"efficiency":[55],"performance.":[58],"key":[60],"idea":[61],"of":[62,68,130,190],"TSM":[63,99,105,120,145,164],"is":[64,195],"to":[65,89,150],"shift":[66],"part":[67],"the":[69,72,111,114,176,188],"channels":[70],"along":[71],"dimension,":[74],"thus":[75],"facilitate":[76],"information":[77],"exchanged":[78],"among":[79],"neighboring":[80],"frames.":[81],"It":[82],"be":[84],"inserted":[85],"into":[86],"modeling":[92],"zero":[94,97],"parameters.":[98],"offers":[100],"several":[101],"unique":[102],"advantages.":[103],"Firstly,":[104],"has":[106,121,146],"performance;":[108],"it":[109,124],"ranks":[110],"first":[112],"on":[113,138,157],"Something-Something":[115],"leaderboard":[116],"upon":[117],"submission.":[118],"Secondly,":[119],"efficiency;":[123],"achieves":[125],"frame":[128],"rate":[129],"74fps":[131],"29fps":[133],"for":[134],"online":[135],"recognition":[137],"Jetson":[139],"Nano":[140],"Galaxy":[142],"Note8.":[143],"Thirdly,":[144],"higher":[147],"scalability":[148],"compared":[149],"networks,":[152],"enabling":[153],"large-scale":[154],"Kinetics":[155],"training":[156,189],"1,536":[158],"GPUs":[159],"15":[161],"minutes.":[162],"Lastly,":[163],"enables":[165],"action":[166,184],"concepts":[167],"learning,":[168],"which":[169],"networks":[171],"model;":[173],"visualize":[175],"category":[177],"attention":[178],"map":[179],"find":[181],"spatial-temporal":[183],"detector":[185],"emerges":[186],"during":[187],"classification":[191],"tasks.":[192],"code":[194],"publicly":[196],"available":[197],"https://github.com/mit-han-lab/temporal-shift-module.":[199]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-08-26T07:47:46.906454","created_date":"2025-10-10T00:00:00"}