{"id":"https://openalex.org/W7154049667","doi":"https://doi.org/10.48550/arxiv.2604.08573","title":"Silhouette Loss: Differentiable Global Structure Learning for Deep Representations","display_name":"Silhouette Loss: Differentiable Global Structure Learning for Deep Representations","publication_year":2026,"publication_date":"2026-03-27","ids":{"openalex":"https://openalex.org/W7154049667","doi":"https://doi.org/10.48550/arxiv.2604.08573"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.08573","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08573","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.2604.08573","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5089225274","display_name":"Matheus Vin\u00edcius Todescato","orcid":"https://orcid.org/0000-0001-7568-8784"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Todescato, Matheus Vin\u00edcius","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5066798979","display_name":"Joel Lu\u00eds Carbonera","orcid":"https://orcid.org/0000-0002-4499-3601"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Carbonera, Joel Lu\u00eds","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":2,"corresponding_author_ids":[],"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/T11448","display_name":"Face recognition and analysis","score":0.7717000246047974,"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/T11448","display_name":"Face recognition and analysis","score":0.7717000246047974,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.0421999990940094,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.03909999877214432,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/silhouette","display_name":"Silhouette","score":0.9049999713897705},{"id":"https://openalex.org/keywords/pairwise-comparison","display_name":"Pairwise comparison","score":0.7657999992370605},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.6272000074386597},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.5907999873161316},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.5771999955177307},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.5297999978065491},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.5184000134468079},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.48919999599456787},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.4643000066280365}],"concepts":[{"id":"https://openalex.org/C58103923","wikidata":"https://www.wikidata.org/wiki/Q2286025","display_name":"Silhouette","level":2,"score":0.9049999713897705},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.7657999992370605},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6323999762535095},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.6272000074386597},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.5907999873161316},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.5771999955177307},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.5297999978065491},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.5184000134468079},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.48919999599456787},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4821000099182129},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4643000066280365},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4498000144958496},{"id":"https://openalex.org/C202615002","wikidata":"https://www.wikidata.org/wiki/Q783507","display_name":"Differentiable function","level":2,"score":0.4287000000476837},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.3871000111103058},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.35929998755455017},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.31859999895095825},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.31630000472068787},{"id":"https://openalex.org/C17212007","wikidata":"https://www.wikidata.org/wiki/Q5511111","display_name":"Fuzzy clustering","level":3,"score":0.30959999561309814},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.28540000319480896},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.2851000130176544},{"id":"https://openalex.org/C164752517","wikidata":"https://www.wikidata.org/wiki/Q5570875","display_name":"Global optimization","level":2,"score":0.2797999978065491},{"id":"https://openalex.org/C105611402","wikidata":"https://www.wikidata.org/wiki/Q2976589","display_name":"Spectral clustering","level":3,"score":0.2777999937534332},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.27320000529289246},{"id":"https://openalex.org/C92835128","wikidata":"https://www.wikidata.org/wiki/Q1277447","display_name":"Hierarchical clustering","level":3,"score":0.26980000734329224},{"id":"https://openalex.org/C89600930","wikidata":"https://www.wikidata.org/wiki/Q1423946","display_name":"Segmentation","level":2,"score":0.26600000262260437},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.25600001215934753},{"id":"https://openalex.org/C64543145","wikidata":"https://www.wikidata.org/wiki/Q162942","display_name":"Intersection (aeronautics)","level":2,"score":0.2531000077724457},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.2513999938964844},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.2500999867916107}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.08573","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08573","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.2604.08573","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.08573","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":[{"score":0.7929377555847168,"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Learning":[0],"discriminative":[1],"representations":[2],"is":[3,140],"a":[4,76,104,149],"central":[5],"goal":[6],"of":[7,107,241],"supervised":[8,44,144],"deep":[9,236],"learning.":[10,146],"While":[11],"cross-entropy":[12],"(CE)":[13,210],"remains":[14],"the":[15,29,82,101,189,197,201],"dominant":[16],"objective":[17,79,151],"for":[18,235],"classification,":[19],"it":[20],"does":[21],"not":[22],"explicitly":[23],"enforce":[24],"desirable":[25],"geometric":[26],"properties":[27],"in":[28,100,247],"embedding":[30],"space,":[31],"such":[32],"as":[33,232],"intra-class":[34],"compactness":[35],"and":[36,48,66,139,160,183,195,211,244],"inter-class":[37],"separation.":[38],"Existing":[39],"metric":[40,185],"learning":[41,46,186],"approaches,":[42],"including":[43],"contrastive":[45,145],"(SupCon)":[47],"proxy-based":[49,59],"methods,":[50],"address":[51],"this":[52,69],"limitation":[53],"by":[54,81],"operating":[55],"on":[56,166],"pairwise":[57,90,158],"or":[58],"relationships,":[60],"but":[61],"often":[62],"increase":[63],"computational":[64,220],"cost":[65],"complexity.":[67],"In":[68],"work,":[70],"we":[71],"introduce":[72],"Soft":[73,130,176],"Silhouette":[74,131,177],"Loss,":[75],"novel":[77],"differentiable":[78,233],"inspired":[80],"classical":[83,226],"silhouette":[84],"coefficient":[85],"from":[86,208],"clustering":[87,227],"analysis.":[88],"Unlike":[89],"objectives,":[91],"our":[92],"formulation":[93,191],"evaluates":[94],"each":[95],"sample":[96],"against":[97],"all":[98],"classes":[99],"batch,":[102],"providing":[103],"batch-level":[105],"notion":[106],"global":[108,161,245],"structure.":[109,163],"The":[110],"proposed":[111],"loss":[112],"directly":[113],"encourages":[114],"samples":[115],"to":[116,119,124,143,214],"be":[117,134,230],"closer":[118],"their":[120],"own":[121],"class":[122],"than":[123],"competing":[125],"classes,":[126],"while":[127,216],"remaining":[128],"lightweight.":[129],"Loss":[132,178],"can":[133,229],"seamlessly":[135],"combined":[136,198],"with":[137,175],"cross-entropy,":[138],"also":[141],"complementary":[142],"We":[147],"propose":[148],"hybrid":[150,190],"that":[152,225],"integrates":[153],"them,":[154],"jointly":[155],"optimizing":[156],"local":[157,243],"consistency":[159],"cluster":[162],"Extensive":[164],"experiments":[165],"seven":[167],"diverse":[168],"datasets":[169],"demonstrate":[170],"that:":[171],"(i)":[172],"augmenting":[173],"CE":[174,182],"consistently":[179],"improves":[180],"over":[181],"other":[184],"baselines;":[187],"(ii)":[188],"outperforms":[192],"SupCon":[193],"alone;":[194],"(iii)":[196],"method":[199],"achieves":[200],"best":[202],"performance,":[203],"improving":[204],"average":[205],"top-1":[206],"accuracy":[207],"36.71%":[209],"37.85%":[212],"(SupCon2)":[213],"39.08%,":[215],"incurring":[217],"substantially":[218],"lower":[219],"overhead.":[221],"These":[222],"results":[223],"suggest":[224],"principles":[228],"reinterpreted":[231],"objectives":[234],"learning,":[237],"enabling":[238],"efficient":[239],"optimization":[240],"both":[242],"structure":[246],"representation":[248],"spaces.":[249]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-04-14T00:00:00"}