{"id":"https://openalex.org/W7127167122","doi":"https://doi.org/10.1145/3787256.3787270","title":"A Highly Area-Efficient Transformer Accelerator for Edge Devices Based on Cross-Shaped Window Attention","display_name":"A Highly Area-Efficient Transformer Accelerator for Edge Devices Based on Cross-Shaped Window Attention","publication_year":2025,"publication_date":"2025-11-21","ids":{"openalex":"https://openalex.org/W7127167122","doi":"https://doi.org/10.1145/3787256.3787270"},"language":null,"primary_location":{"id":"doi:10.1145/3787256.3787270","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3787256.3787270","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3787256.3787270?download=true","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2025 8th International Conference on Computational Intelligence and Intelligent Systems","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3787256.3787270?download=true","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5124852116","display_name":"Yue-Hang Li","orcid":null},"institutions":[{"id":"https://openalex.org/I118292597","display_name":"National Taipei University of Technology","ror":"https://ror.org/00cn92c09","country_code":"TW","type":"education","lineage":["https://openalex.org/I118292597"]}],"countries":["TW"],"is_corresponding":true,"raw_author_name":"Yue-Hang Li","raw_affiliation_strings":["National Taipei University of Technology, Taipei, Taiwan,"],"raw_orcid":"https://orcid.org/0009-0003-3903-1410","affiliations":[{"raw_affiliation_string":"National Taipei University of Technology, Taipei, Taiwan,","institution_ids":["https://openalex.org/I118292597"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124796869","display_name":"Tzu-Lun Fang","orcid":null},"institutions":[{"id":"https://openalex.org/I118292597","display_name":"National Taipei University of Technology","ror":"https://ror.org/00cn92c09","country_code":"TW","type":"education","lineage":["https://openalex.org/I118292597"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Tzu-Lun Fang","raw_affiliation_strings":["National Taipei University of Technology, Taipei, Taiwan,"],"raw_orcid":"https://orcid.org/0009-0001-3855-1655","affiliations":[{"raw_affiliation_string":"National Taipei University of Technology, Taipei, Taiwan,","institution_ids":["https://openalex.org/I118292597"]}]},{"author_position":"last","author":{"id":null,"display_name":"Yu-Cheng Fan","orcid":"https://orcid.org/0000-0002-9599-6415"},"institutions":[{"id":"https://openalex.org/I118292597","display_name":"National Taipei University of Technology","ror":"https://ror.org/00cn92c09","country_code":"TW","type":"education","lineage":["https://openalex.org/I118292597"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Yu-Cheng Fan","raw_affiliation_strings":["National Taipei University of Technology, Taipei, Taiwan,"],"raw_orcid":"https://orcid.org/0000-0002-9599-6415","affiliations":[{"raw_affiliation_string":"National Taipei University of Technology, Taipei, Taiwan,","institution_ids":["https://openalex.org/I118292597"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5124852116"],"corresponding_institution_ids":["https://openalex.org/I118292597"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.60392321,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"96","last_page":"100"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11992","display_name":"CCD and CMOS Imaging Sensors","score":0.6984999775886536,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11992","display_name":"CCD and CMOS Imaging Sensors","score":0.6984999775886536,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.06530000269412994,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.0478999987244606,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/modular-design","display_name":"Modular design","score":0.49239999055862427},{"id":"https://openalex.org/keywords/hardware-acceleration","display_name":"Hardware acceleration","score":0.4602000117301941},{"id":"https://openalex.org/keywords/computational-complexity-theory","display_name":"Computational complexity theory","score":0.4564000070095062},{"id":"https://openalex.org/keywords/edge-device","display_name":"Edge device","score":0.44760000705718994},{"id":"https://openalex.org/keywords/datapath","display_name":"Datapath","score":0.3935999870300293},{"id":"https://openalex.org/keywords/softmax-function","display_name":"Softmax function","score":0.3813000023365021},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.37450000643730164},{"id":"https://openalex.org/keywords/chip","display_name":"Chip","score":0.3723999857902527},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.35679998993873596}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7143999934196472},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.4968999922275543},{"id":"https://openalex.org/C101468663","wikidata":"https://www.wikidata.org/wiki/Q1620158","display_name":"Modular design","level":2,"score":0.49239999055862427},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.4602000117301941},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.4564000070095062},{"id":"https://openalex.org/C138236772","wikidata":"https://www.wikidata.org/wiki/Q25098575","display_name":"Edge device","level":3,"score":0.44760000705718994},{"id":"https://openalex.org/C2781198647","wikidata":"https://www.wikidata.org/wiki/Q1633673","display_name":"Datapath","level":2,"score":0.3935999870300293},{"id":"https://openalex.org/C188441871","wikidata":"https://www.wikidata.org/wiki/Q7554146","display_name":"Softmax function","level":3,"score":0.3813000023365021},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.37450000643730164},{"id":"https://openalex.org/C165005293","wikidata":"https://www.wikidata.org/wiki/Q1074500","display_name":"Chip","level":2,"score":0.3723999857902527},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.35679998993873596},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.35190001130104065},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.3481000065803528},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.34439998865127563},{"id":"https://openalex.org/C103088060","wikidata":"https://www.wikidata.org/wiki/Q1062839","display_name":"Error detection and correction","level":2,"score":0.33390000462532043},{"id":"https://openalex.org/C172385210","wikidata":"https://www.wikidata.org/wiki/Q5339","display_name":"Transistor","level":3,"score":0.32100000977516174},{"id":"https://openalex.org/C164660894","wikidata":"https://www.wikidata.org/wiki/Q2037833","display_name":"Piecewise","level":2,"score":0.3125999867916107},{"id":"https://openalex.org/C2778456923","wikidata":"https://www.wikidata.org/wiki/Q5337692","display_name":"Edge computing","level":3,"score":0.30869999527931213},{"id":"https://openalex.org/C65232700","wikidata":"https://www.wikidata.org/wiki/Q5656403","display_name":"Hardware architecture","level":3,"score":0.2915000021457672},{"id":"https://openalex.org/C530198007","wikidata":"https://www.wikidata.org/wiki/Q80831","display_name":"Integrated circuit","level":2,"score":0.2915000021457672},{"id":"https://openalex.org/C118021083","wikidata":"https://www.wikidata.org/wiki/Q610398","display_name":"System on a chip","level":2,"score":0.29030001163482666},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.27970001101493835},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.2791999876499176},{"id":"https://openalex.org/C90702460","wikidata":"https://www.wikidata.org/wiki/Q1055112","display_name":"Circuit complexity","level":3,"score":0.274399995803833},{"id":"https://openalex.org/C24326235","wikidata":"https://www.wikidata.org/wiki/Q126095","display_name":"Electronic engineering","level":1,"score":0.26460000872612},{"id":"https://openalex.org/C17095337","wikidata":"https://www.wikidata.org/wiki/Q2375229","display_name":"Piecewise linear function","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C135402231","wikidata":"https://www.wikidata.org/wiki/Q898440","display_name":"Dissipation","level":2,"score":0.25429999828338623}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3787256.3787270","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3787256.3787270","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3787256.3787270?download=true","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2025 8th International Conference on Computational Intelligence and Intelligent Systems","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3787256.3787270","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3787256.3787270","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3787256.3787270?download=true","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2025 8th International Conference on Computational Intelligence and Intelligent Systems","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy","score":0.8994458317756653}],"awards":[],"funders":[{"id":"https://openalex.org/F4320329659","display_name":"Thailand Science Research and Innovation","ror":null},{"id":"https://openalex.org/F4320331164","display_name":"National Science and Technology Council","ror":"https://ror.org/00wnb9798"}],"has_content":{"pdf":true,"grobid_xml":false},"content_urls":{"pdf":"https://content.openalex.org/works/W7127167122.pdf"},"referenced_works_count":3,"referenced_works":["https://openalex.org/W2169471437","https://openalex.org/W2173365639","https://openalex.org/W4224272693"],"related_works":[],"abstract_inverted_index":{"This":[0],"study":[1,49,170],"proposes":[2],"a":[3,20,51,75,81,137,172],"highly":[4],"area-efficient":[5],"Vision":[6],"Transformer":[7,155],"hardware":[8,104,112,120,165],"accelerator,":[9],"particularly":[10],"suitable":[11],"for":[12,160],"edge":[13,161],"computing":[14],"environments.":[15],"The":[16,147],"proposed":[17,111,167],"design":[18],"incorporates":[19],"cross-window":[21],"attention":[22],"mechanism":[23],"and":[24,38,63,103,122,135,177],"block":[25],"matrix":[26,36],"operations":[27],"to":[28,59],"effectively":[29],"reduce":[30,64,101],"the":[31,70,86,91,110,154,164],"computational":[32,40,138],"complexity":[33],"of":[34,132,140],"large":[35],"multiplications":[37],"improve":[39,60],"efficiency.":[41,126,179],"To":[42],"address":[43],"internal":[44],"chip":[45,133],"memory":[46,55],"limitations,":[47],"this":[48,169],"devised":[50],"refined":[52],"static":[53],"random-access":[54],"(SRAM)":[56],"allocation":[57],"strategy":[58],"data":[61],"reuse":[62],"overall":[65],"area":[66,125,134],"requirements.":[67],"In":[68],"addition,":[69],"Softmax":[71],"operation":[72],"module":[73],"adopts":[74],"linear":[76],"piecewise":[77],"approximation":[78],"combined":[79],"with":[80,85],"log2-based":[82],"function.":[83],"Compared":[84],"traditional":[87],"exponential":[88],"function":[89],"implementation,":[90],"average":[92],"error":[93],"is":[94],"only":[95,129],"about":[96],"2.7%,":[97],"which":[98],"can":[99],"significantly":[100,143],"latency":[102],"costs.":[105],"Experimental":[106],"results":[107],"show":[108],"that":[109],"accelerator":[113,166],"maintains":[114],"competitive":[115],"classification":[116],"performance":[117],"under":[118],"strict":[119],"constraints":[121],"exhibits":[123],"excellent":[124],"It":[127],"requires":[128],"1.79":[130],"mm\u00b2":[131],"achieves":[136,171],"efficiency":[139],"253.13":[141],"GOPS/mm\u00b2,":[142],"outperforming":[144],"existing":[145],"solutions.":[146],"modular":[148],"architecture":[149],"also":[150],"facilitates":[151],"integration":[152],"into":[153],"model,":[156],"demonstrating":[157],"indicating":[158],"competitiveness":[159],"deployment.":[162],"Overall,":[163],"in":[168],"balance":[173],"between":[174],"accuracy,":[175],"performance,":[176],"energy":[178]},"counts_by_year":[],"updated_date":"2026-03-11T06:11:40.159057","created_date":"2026-02-03T00:00:00"}