{"id":"https://openalex.org/W7154393768","doi":"https://doi.org/10.48550/arxiv.2604.11564","title":"Training-Free Model Ensemble for Single-Image Super-Resolution via Strong-Branch Compensation","display_name":"Training-Free Model Ensemble for Single-Image Super-Resolution via Strong-Branch Compensation","publication_year":2026,"publication_date":"2026-04-13","ids":{"openalex":"https://openalex.org/W7154393768","doi":"https://doi.org/10.48550/arxiv.2604.11564"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.11564","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.11564","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.11564","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133618567","display_name":"Gengjia Chang","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Chang, Gengjia","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133582953","display_name":"Xining Ge","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ge, Xining","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102781181","display_name":"Weijun Yuan","orcid":"https://orcid.org/0009-0003-3215-4848"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuan, Weijun","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133619480","display_name":"Zhan Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Zhan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Song, Qiurong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Qiurong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076561018","display_name":"Luen Zhu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Luen","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133601212","display_name":"Shuhong Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Shuhong","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5133618567"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"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/T11105","display_name":"Advanced Image Processing Techniques","score":0.9648000001907349,"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/T11105","display_name":"Advanced Image Processing Techniques","score":0.9648000001907349,"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/T11165","display_name":"Image and Video Quality Assessment","score":0.005200000014156103,"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/T11659","display_name":"Advanced Image Fusion Techniques","score":0.005100000184029341,"subfield":{"id":"https://openalex.org/subfields/2214","display_name":"Media Technology"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4587000012397766},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.4535999894142151},{"id":"https://openalex.org/keywords/upgrade","display_name":"Upgrade","score":0.453000009059906},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.42649999260902405},{"id":"https://openalex.org/keywords/boosting","display_name":"Boosting (machine learning)","score":0.4074000120162964},{"id":"https://openalex.org/keywords/constraint","display_name":"Constraint (computer-aided design)","score":0.39739999175071716},{"id":"https://openalex.org/keywords/compensation","display_name":"Compensation (psychology)","score":0.38600000739097595},{"id":"https://openalex.org/keywords/bilinear-interpolation","display_name":"Bilinear interpolation","score":0.3767000138759613}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5960999727249146},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4587000012397766},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.4535999894142151},{"id":"https://openalex.org/C2780615140","wikidata":"https://www.wikidata.org/wiki/Q920419","display_name":"Upgrade","level":2,"score":0.453000009059906},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4311000108718872},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.42649999260902405},{"id":"https://openalex.org/C46686674","wikidata":"https://www.wikidata.org/wiki/Q466303","display_name":"Boosting (machine learning)","level":2,"score":0.4074000120162964},{"id":"https://openalex.org/C2776036281","wikidata":"https://www.wikidata.org/wiki/Q48769818","display_name":"Constraint (computer-aided design)","level":2,"score":0.39739999175071716},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.39309999346733093},{"id":"https://openalex.org/C2780023022","wikidata":"https://www.wikidata.org/wiki/Q1338171","display_name":"Compensation (psychology)","level":2,"score":0.38600000739097595},{"id":"https://openalex.org/C205203396","wikidata":"https://www.wikidata.org/wiki/Q612143","display_name":"Bilinear interpolation","level":2,"score":0.3767000138759613},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.3763999938964844},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.35600000619888306},{"id":"https://openalex.org/C157202957","wikidata":"https://www.wikidata.org/wiki/Q1659609","display_name":"Image warping","level":2,"score":0.32249999046325684},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3181000053882599},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.3149999976158142},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.3124000132083893},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.31150001287460327},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.2957000136375427},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.27549999952316284},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.2741999924182892},{"id":"https://openalex.org/C49608258","wikidata":"https://www.wikidata.org/wiki/Q611705","display_name":"Bicubic interpolation","level":4,"score":0.27230000495910645},{"id":"https://openalex.org/C45493050","wikidata":"https://www.wikidata.org/wiki/Q7884934","display_name":"Unified Model","level":2,"score":0.26739999651908875},{"id":"https://openalex.org/C45347329","wikidata":"https://www.wikidata.org/wiki/Q5166604","display_name":"Convolution (computer science)","level":3,"score":0.25029999017715454}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.11564","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.11564","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.11564","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.11564","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[{"display_name":"Industry, innovation and infrastructure","id":"https://metadata.un.org/sdg/9","score":0.4488288164138794}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Single-image":[0],"super-resolution":[1,204],"has":[2],"progressed":[3],"from":[4],"deep":[5],"convolutional":[6],"baselines":[7],"to":[8,147],"stronger":[9],"Transformer":[10],"and":[11,28,46,121,165,197],"state-space":[12],"architectures,":[13],"yet":[14],"the":[15,47,116,125,148,156,162,168,175],"corresponding":[16],"performance":[17],"gains":[18],"typically":[19],"come":[20],"with":[21,39,91,102],"higher":[22],"training":[23],"cost,":[24],"longer":[25,52],"engineering":[26],"iteration,":[27],"heavier":[29],"deployment":[30],"burden.":[31],"In":[32],"many":[33],"practical":[34],"settings,":[35],"multiple":[36],"pretrained":[37],"models":[38],"partially":[40],"complementary":[41],"behaviors":[42],"are":[43,122],"already":[44],"available,":[45],"binding":[48],"constraint":[49],"is":[50,83],"no":[51],"architectural":[53,70],"capacity":[54],"but":[55],"how":[56],"effectively":[57],"their":[58],"outputs":[59],"can":[60],"be":[61],"combined":[62],"without":[63,133],"additional":[64,141],"training.":[65],"Rather":[66],"than":[67],"pursuing":[68],"further":[69],"redesign,":[71],"this":[72],"paper":[73],"proposes":[74],"a":[75,87,99,129,180,195],"training-free":[76],"output-level":[77,192],"ensemble":[78],"framework.":[79],"A":[80],"dual-branch":[81],"pipeline":[82],"constructed":[84],"in":[85,124,172],"which":[86],"Hybrid":[88],"attention":[89],"network":[90],"TLC":[92],"inference":[93],"provides":[94,194],"stable":[95],"main":[96],"reconstruction,":[97],"while":[98],"MambaIRv2":[100],"branch":[101,164,171],"geometric":[103],"self-ensemble":[104],"supplies":[105],"strong":[106,170],"compensation":[107,193],"for":[108,202],"high-frequency":[109],"detail":[110],"recovery.":[111],"The":[112],"two":[113],"branches":[114],"process":[115],"same":[117],"low-resolution":[118],"input":[119],"independently":[120],"fused":[123],"image":[126],"space":[127],"via":[128],"lightweight":[130],"weighted":[131],"combination,":[132],"updating":[134],"any":[135],"model":[136],"parameters":[137],"or":[138],"introducing":[139],"an":[140],"trainable":[142],"module.":[143],"As":[144],"our":[145],"solution":[146],"NTIRE":[149],"2026":[150],"Image":[151],"Super-Resolution":[152],"($\\times":[153],"4$)":[154],"Challenge,":[155],"proposed":[157],"design":[158],"consistently":[159],"improves":[160],"over":[161],"base":[163],"slightly":[166],"exceeds":[167],"pure":[169],"PSNR":[173],"at":[174],"best":[176],"operating":[177],"point":[178],"under":[179],"unified":[181],"DIV2K":[182],"bicubic":[183],"$\\times":[184],"4$":[185],"evaluation":[186],"protocol.":[187],"Ablation":[188],"studies":[189],"confirm":[190],"that":[191],"low-overhead":[196],"practically":[198],"accessible":[199],"upgrade":[200],"path":[201],"existing":[203],"systems.":[205]},"counts_by_year":[],"updated_date":"2026-04-21T08:09:41.155169","created_date":"2026-04-15T00:00:00"}