{"id":"https://openalex.org/W3095592098","doi":null,"title":"What's in a Loss Function for Image Classification?","display_name":"What's in a Loss Function for Image Classification?","publication_year":2020,"publication_date":"2020-10-30","ids":{"openalex":"https://openalex.org/W3095592098","mag":"3095592098"},"language":"en","primary_location":{"id":"mag:3095592098","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2010.16402","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2010.16402","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5042760897","display_name":"Simon Kornblith","orcid":"https://orcid.org/0000-0002-9088-2443","observed_orcids":["https://orcid.org/0000-0002-9088-2443"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Simon Kornblith","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0002-9088-2443","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108652283","display_name":"Honglak Lee","orcid":"https://orcid.org/0000-0002-4109-327X","observed_orcids":["https://orcid.org/0000-0002-4109-327X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Honglak Lee","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100443205","display_name":"Ting Chen","orcid":"https://orcid.org/0000-0002-9599-871X","observed_orcids":["https://orcid.org/0000-0002-9599-871X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ting Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5103947107","display_name":"Mohammad Norouzi","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mohammad Norouzi","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":7,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.5375999808311462,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.5375999808311462,"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":0.3328000009059906,"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"}},{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.12120000272989273,"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/loss-functions","display_name":"loss functions","score":0.8799999952316284},{"id":"https://openalex.org/keywords/label-smoothing","display_name":"label smoothing","score":0.9559999704360962},{"id":"https://openalex.org/keywords/dropout","display_name":"dropout","score":0.7549999952316284},{"id":"https://openalex.org/keywords/image-classification","display_name":"image classification","score":0.9120000004768372},{"id":"https://openalex.org/keywords/model-calibration","display_name":"model calibration","score":0.7329999804496765},{"id":"https://openalex.org/keywords/out-of-distribution-robustness","display_name":"out-of-distribution robustness","score":0.9990000128746033},{"id":"https://openalex.org/keywords/representational-similarity","display_name":"representational similarity","score":0.9710000157356262},{"id":"https://openalex.org/keywords/class-separation","display_name":"class separation","score":0.6899999976158142},{"id":"https://openalex.org/keywords/transfer-learning","display_name":"transfer learning","score":0.6620000004768372}],"concepts":[{"id":"https://openalex.org/C188441871","wikidata":"https://www.wikidata.org/wiki/Q7554146","display_name":"Softmax function","level":3,"score":0.9745200276374817},{"id":"https://openalex.org/C3770464","wikidata":"https://www.wikidata.org/wiki/Q775963","display_name":"Smoothing","level":2,"score":0.7505029439926147},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6482874155044556},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6123257279396057},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.5508424043655396},{"id":"https://openalex.org/C167981619","wikidata":"https://www.wikidata.org/wiki/Q1685498","display_name":"Cross entropy","level":3,"score":0.5009374618530273},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4838827848434448},{"id":"https://openalex.org/C75294576","wikidata":"https://www.wikidata.org/wiki/Q5165192","display_name":"Contextual image classification","level":3,"score":0.45795735716819763},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.4543645977973938},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4278549551963806},{"id":"https://openalex.org/C106301342","wikidata":"https://www.wikidata.org/wiki/Q4117933","display_name":"Entropy (arrow of time)","level":2,"score":0.4276210069656372},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.42385995388031006},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.25707054138183594},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.07995060086250305},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C55493867","wikidata":"https://www.wikidata.org/wiki/Q7094","display_name":"Biochemistry","level":1,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C104317684","wikidata":"https://www.wikidata.org/wiki/Q7187","display_name":"Gene","level":2,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:3095592098","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2010.16402","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null}],"best_oa_location":{"id":"mag:3095592098","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2010.16402","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":38,"referenced_works":["https://openalex.org/W183625566","https://openalex.org/W1977295328","https://openalex.org/W2017814585","https://openalex.org/W2123045220","https://openalex.org/W2140766383","https://openalex.org/W2142387771","https://openalex.org/W2152722485","https://openalex.org/W2194775991","https://openalex.org/W2593634001","https://openalex.org/W2600537992","https://openalex.org/W2601450892","https://openalex.org/W2622263826","https://openalex.org/W2626967530","https://openalex.org/W2765407302","https://openalex.org/W2785885194","https://openalex.org/W2804935296","https://openalex.org/W2898737887","https://openalex.org/W2942810103","https://openalex.org/W2947707615","https://openalex.org/W2948643221","https://openalex.org/W2950703532","https://openalex.org/W2952502547","https://openalex.org/W2962876041","https://openalex.org/W2963060032","https://openalex.org/W2963466847","https://openalex.org/W2963697299","https://openalex.org/W2964273174","https://openalex.org/W2964292098","https://openalex.org/W2970206392","https://openalex.org/W2980149079","https://openalex.org/W3004639598","https://openalex.org/W3009419078","https://openalex.org/W3011927872","https://openalex.org/W3018378048","https://openalex.org/W3034363135","https://openalex.org/W3034893234","https://openalex.org/W3037492894","https://openalex.org/W3177096435"],"related_works":["https://openalex.org/W3131044704","https://openalex.org/W3210131709","https://openalex.org/W3125317419","https://openalex.org/W4402774047","https://openalex.org/W2885023989","https://openalex.org/W6967281039","https://openalex.org/W2963689346","https://openalex.org/W4390045702","https://openalex.org/W3134134808","https://openalex.org/W2911303748"],"abstract_inverted_index":{"It":[0],"is":[1,21],"common":[2],"to":[3,9,23,45,96,113,155],"use":[4],"the":[5,97,114,132,136,145],"softmax":[6,33,125],"cross-entropy":[7,34],"loss":[8,54,126],"train":[10],"neural":[11],"networks":[12,78,89],"on":[13,61,144],"classification":[14,63],"datasets":[15],"where":[16],"a":[17,51],"single":[18],"class":[19,129],"label":[20,36],"assigned":[22],"each":[24],"example.":[25],"However,":[26,83],"it":[27],"has":[28],"been":[29],"shown":[30],"that":[31,109,118,121,152],"modifying":[32],"with":[35,80,91],"smoothing":[37],"or":[38],"regularizers":[39],"such":[40],"as":[41],"dropout":[42],"can":[43],"lead":[44],"higher":[46],"performance.":[47],"This":[48],"paper":[49],"studies":[50],"variety":[52],"of":[53,88,135],"functions":[55],"and":[56,74],"output":[57],"layer":[58,134],"regularization":[59],"strategies":[60],"image":[62],"tasks.":[64,157],"We":[65,116],"observe":[66],"meaningful":[67],"differences":[68,84,105],"in":[69,85,131,150],"model":[70],"predictions,":[71],"accuracy,":[72],"calibration,":[73],"out-of-distribution":[75],"robustness":[76],"for":[77,141],"trained":[79,90],"different":[81,92],"objectives.":[82],"hidden":[86],"representations":[87],"objectives":[93,120],"are":[94,110],"restricted":[95],"last":[98],"few":[99],"layers;":[100],"representational":[101],"similarity":[102],"reveals":[103],"no":[104],"among":[106],"network":[107],"layers":[108],"not":[111],"close":[112],"output.":[115],"show":[117],"all":[119],"improve":[122],"over":[123],"vanilla":[124],"produce":[127],"greater":[128],"separation":[130],"penultimate":[133],"network,":[137],"which":[138],"potentially":[139],"accounts":[140],"improved":[142],"performance":[143],"original":[146],"task,":[147],"but":[148],"results":[149],"features":[151],"transfer":[153],"worse":[154],"other":[156]},"counts_by_year":[{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":4}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2020-11-09T00:00:00"}