{"id":"https://openalex.org/W3094869948","doi":null,"title":"Understanding when spatial transformer networks do not support invariance, and what to do about it","display_name":"Understanding when spatial transformer networks do not support invariance, and what to do about it","publication_year":2020,"publication_date":"2020-01-01","ids":{"openalex":"https://openalex.org/W3094869948","mag":"3094869948"},"language":"en","primary_location":{"id":"pmh:oai:DiVA.org:kth-272971","doi":null,"is_oa":true,"landing_page_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-272971","pdf_url":null,"source":{"id":"https://openalex.org/S4306401559","display_name":"KTH Publication Database DiVA (KTH Royal Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I86987016","host_organization_name":"KTH Royal Institute of Technology","host_organization_lineage":["https://openalex.org/I86987016"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/report"},"volume":null,"type":"report","is_traditional":false,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-272971","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5031910861","display_name":"Lukas Finnveden","orcid":"https://orcid.org/0000-0001-8548-5788","observed_orcids":["https://orcid.org/0000-0001-8548-5788"]},"institutions":[{"id":"https://openalex.org/I86987016","display_name":"KTH Royal Institute of Technology","ror":"https://ror.org/026vcq606","country_code":"SE","type":"education","lineage":["https://openalex.org/I86987016"]}],"countries":["SE"],"is_corresponding":false,"raw_author_name":"Finnveden, Lukas","raw_affiliation_strings":["Division of Computational Science and Technology, Computational Brain Science Lab, KTH Royal Institute of Technology, Stockholm, Sweden"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Division of Computational Science and Technology, Computational Brain Science Lab, KTH Royal Institute of Technology, Stockholm, Sweden","institution_ids":["https://openalex.org/I86987016"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072246179","display_name":"Ylva Jansson","orcid":"https://orcid.org/0000-0003-0011-6444","observed_orcids":["https://orcid.org/0000-0003-0011-6444"]},"institutions":[{"id":"https://openalex.org/I86987016","display_name":"KTH Royal Institute of Technology","ror":"https://ror.org/026vcq606","country_code":"SE","type":"education","lineage":["https://openalex.org/I86987016"]}],"countries":["SE"],"is_corresponding":false,"raw_author_name":"Jansson, Ylva","raw_affiliation_strings":["Division of Computational Science and Technology, Computational Brain Science Lab, KTH Royal Institute of Technology, Stockholm, Sweden"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Division of Computational Science and Technology, Computational Brain Science Lab, KTH Royal Institute of Technology, Stockholm, Sweden","institution_ids":["https://openalex.org/I86987016"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5054396186","display_name":"Tony Lindeberg","orcid":"https://orcid.org/0000-0002-9081-2170","observed_orcids":["https://orcid.org/0000-0002-9081-2170"]},"institutions":[{"id":"https://openalex.org/I86987016","display_name":"KTH Royal Institute of Technology","ror":"https://ror.org/026vcq606","country_code":"SE","type":"education","lineage":["https://openalex.org/I86987016"]}],"countries":["SE"],"is_corresponding":false,"raw_author_name":"Lindeberg, Tony","raw_affiliation_strings":["Division of Computational Science and Technology, Computational Brain Science Lab, KTH Royal Institute of Technology, Stockholm, Sweden"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Division of Computational Science and Technology, Computational Brain Science Lab, KTH Royal Institute of Technology, Stockholm, Sweden","institution_ids":["https://openalex.org/I86987016"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I86987016"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":21,"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":0.8999000191688538,"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":0.8999000191688538,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.06120000034570694,"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":0.02810000069439411,"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/spatial-transformer-network","display_name":"spatial transformer network","score":1.0},{"id":"https://openalex.org/keywords/invariance","display_name":"invariance","score":0.7350000143051147},{"id":"https://openalex.org/keywords/image-transformation","display_name":"image transformation","score":0.6269999742507935},{"id":"https://openalex.org/keywords/weight-sharing","display_name":"parameter sharing","score":0.5979999899864197},{"id":"https://openalex.org/keywords/classification-accuracy","display_name":"classification accuracy","score":0.6460000276565552}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7341403365135193},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.6663104295730591},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6466312408447266},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.6117292642593384},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.5169591307640076},{"id":"https://openalex.org/C204241405","wikidata":"https://www.wikidata.org/wiki/Q461499","display_name":"Transformation (genetics)","level":3,"score":0.5162970423698425},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.47386980056762695},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.4582684338092804},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.41783392429351807},{"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},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.0},{"id":"https://openalex.org/C55493867","wikidata":"https://www.wikidata.org/wiki/Q7094","display_name":"Biochemistry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:DiVA.org:kth-272971","doi":null,"is_oa":true,"landing_page_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-272971","pdf_url":null,"source":{"id":"https://openalex.org/S4306401559","display_name":"KTH Publication Database DiVA (KTH Royal Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I86987016","host_organization_name":"KTH Royal Institute of Technology","host_organization_lineage":["https://openalex.org/I86987016"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/report"}],"best_oa_location":{"id":"pmh:oai:DiVA.org:kth-272971","doi":null,"is_oa":true,"landing_page_url":"http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-272971","pdf_url":null,"source":{"id":"https://openalex.org/S4306401559","display_name":"KTH Publication Database DiVA (KTH Royal Institute of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I86987016","host_organization_name":"KTH Royal Institute of Technology","host_organization_lineage":["https://openalex.org/I86987016"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/report"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G6250578322","display_name":"Scale-space theory for covariant and invariant visual perception","funder_award_id":"2018-03586","funder_id":"https://openalex.org/F4320322581","funder_display_name":"Vetenskapsr\u00e5det"}],"funders":[{"id":"https://openalex.org/F4320322581","display_name":"Vetenskapsr\u00e5det","ror":"https://ror.org/03zttf063"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":31,"referenced_works":["https://openalex.org/W603908379","https://openalex.org/W1237598256","https://openalex.org/W1912570122","https://openalex.org/W1938714998","https://openalex.org/W1970269179","https://openalex.org/W1980911747","https://openalex.org/W2118877769","https://openalex.org/W2165497495","https://openalex.org/W2167383966","https://openalex.org/W2172188317","https://openalex.org/W2179352600","https://openalex.org/W2279221249","https://openalex.org/W2286929393","https://openalex.org/W2335728318","https://openalex.org/W2435623039","https://openalex.org/W2562066862","https://openalex.org/W2576989276","https://openalex.org/W2601564443","https://openalex.org/W2785397276","https://openalex.org/W2792806930","https://openalex.org/W2891202958","https://openalex.org/W2912259665","https://openalex.org/W2950974964","https://openalex.org/W2963383990","https://openalex.org/W2965279912","https://openalex.org/W2970756352","https://openalex.org/W2981512247","https://openalex.org/W3023769537","https://openalex.org/W4231985393","https://openalex.org/W4297688279","https://openalex.org/W4297805934"],"related_works":["https://openalex.org/W3000016346","https://openalex.org/W2901633229","https://openalex.org/W7071594545","https://openalex.org/W4415088028","https://openalex.org/W2562066862","https://openalex.org/W2951005624","https://openalex.org/W2866844117","https://openalex.org/W4287816707","https://openalex.org/W3023769537","https://openalex.org/W4214491563"],"abstract_inverted_index":{"Spatial":[0],"transformer":[1],"networks":[2,10,128,134],"(STNs)":[3],"were":[4,19],"designed":[5],"to":[6,12,15,22,62,80,131],"enable":[7],"convolutional":[8],"neural":[9],"(CNNs)":[11],"learn":[13],"invariance":[14,82],"image":[16,70,168],"transformations.":[17],"STNs":[18,46,76],"originally":[20],"proposed":[21],"transform":[23],"CNN":[24,85],"feature":[25,65,86],"maps":[26,66],"as":[27,29,144],"well":[28],"input":[30],"images.":[31],"This":[32],"enables":[33],"the":[34,56,60,64,97,139,160],"use":[35,118],"of":[36,67,73,119],"more":[37],"complex":[38,120],"features":[39],"when":[40,83],"predicting":[41],"transformation":[42],"parameters.":[43],"However,":[44],"since":[45],"perform":[47],"a":[48,68,90],"purely":[49],"spatial":[50],"transformation,":[51],"they":[52,145],"do":[53],"not,":[54],"in":[55],"general":[57],"case,":[58],"have":[59],"ability":[61],"align":[63],"transformed":[69],"with":[71,106,138],"those":[72],"its":[74],"original.":[75],"are":[77,129],"therefore":[78,111],"unable":[79],"support":[81],"transforming":[84],"maps.":[87],"We":[88,110,122],"present":[89],"simple":[91],"proof":[92],"for":[93,150],"this":[94,102],"and":[95,166],"study":[96],"practical":[98],"implications,":[99],"showing":[100],"that":[101,116,124,135],"inability":[103],"is":[104],"coupled":[105],"decreased":[107],"classification":[108,140,152],"accuracy.":[109],"investigate":[112],"alternative":[113],"STN":[114],"architectures":[115],"make":[117],"features.":[121],"find":[123],"while":[125],"deeper":[126],"localization":[127,133,163],"difficult":[130,155],"train,":[132],"share":[136],"parameters":[137],"network":[141,164],"remain":[142],"stable":[143],"grow":[146],"deeper,":[147],"which":[148],"allows":[149],"higher":[151],"accuracy":[153],"on":[154],"datasets.":[156],"Finally,":[157],"we":[158],"explore":[159],"interaction":[161],"between":[162],"complexity":[165],"iterative":[167],"alignment.":[169]},"counts_by_year":[{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":5},{"year":2022,"cited_by_count":1},{"year":2021,"cited_by_count":6},{"year":2020,"cited_by_count":2}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2020-11-09T00:00:00"}