{"id":"https://openalex.org/W4414902808","doi":"https://doi.org/10.48550/arxiv.2503.15141","title":"Object-Centric Pretraining via Target Encoder Bootstrapping","display_name":"Object-Centric Pretraining via Target Encoder Bootstrapping","publication_year":2025,"publication_date":"2025-03-19","ids":{"openalex":"https://openalex.org/W4414902808","doi":"https://doi.org/10.48550/arxiv.2503.15141"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2503.15141","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2503.15141","pdf_url":"https://arxiv.org/pdf/2503.15141","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":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2503.15141","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"\u0110uki\u0107, Nikola","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"\u0110uki\u0107, Nikola","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067774307","display_name":"Tim Lebailly","orcid":"https://orcid.org/0000-0002-9814-7531","observed_orcids":["https://orcid.org/0000-0002-9814-7531"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lebailly, Tim","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5074816094","display_name":"Tinne Tuytelaars","orcid":"https://orcid.org/0000-0003-3307-9723","observed_orcids":["https://orcid.org/0000-0003-3307-9723"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tuytelaars, Tinne","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":true,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.5303000211715698,"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"}},"topics":[{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.5303000211715698,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.3481000065803528,"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/T11605","display_name":"Visual Attention and Saliency Detection","score":0.04360000044107437,"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/slot-attention","display_name":"slot attention","score":0.7549999952316284},{"id":"https://openalex.org/keywords/self-distillation","display_name":"self-distillation","score":0.9300000071525574},{"id":"https://openalex.org/keywords/exponentially-weighted-moving-average","display_name":"exponential moving average","score":0.7059999704360962},{"id":"https://openalex.org/keywords/unsupervised-object-discovery","display_name":"unsupervised object discovery","score":0.8100000023841858},{"id":"https://openalex.org/keywords/inductive-bias","display_name":"inductive bias","score":0.5989999771118164}],"concepts":[{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.8669999837875366},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7791000008583069},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.755299985408783},{"id":"https://openalex.org/C207609745","wikidata":"https://www.wikidata.org/wiki/Q4944086","display_name":"Bootstrapping (finance)","level":2,"score":0.619700014591217},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.616100013256073},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.5069000124931335},{"id":"https://openalex.org/C2781238097","wikidata":"https://www.wikidata.org/wiki/Q175026","display_name":"Object (grammar)","level":2,"score":0.5062999725341797},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4634999930858612},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4526999890804291},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.42820000648498535},{"id":"https://openalex.org/C77553402","wikidata":"https://www.wikidata.org/wiki/Q13222579","display_name":"Upper and lower bounds","level":2,"score":0.3702999949455261},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3587999939918518},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.3531999886035919},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.3366999924182892},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32170000672340393},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.296099990606308},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.28360000252723694},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.2784000039100647},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.25130000710487366}],"mesh":[],"locations_count":3,"locations":[{"id":"pmh:oai:arXiv.org:2503.15141","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2503.15141","pdf_url":"https://arxiv.org/pdf/2503.15141","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":null,"raw_type":"text"},{"id":"pmh:oai:lirias2repo.kuleuven.be:20.500.12942/765699","doi":null,"is_oa":true,"landing_page_url":"https://lirias.kuleuven.be/handle/20.500.12942/765699","pdf_url":"https://lirias.kuleuven.be/retrieve/fdf47774-6306-495e-b3c8-fa5933182a40","source":{"id":"https://openalex.org/S7407055369","display_name":"Lirias","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"ICLR 2025, Singapore, 24-28 April 2025","raw_type":"info:eu-repo/semantics/conferenceObject"},{"id":"doi:10.48550/arxiv.2503.15141","doi":"https://doi.org/10.48550/arxiv.2503.15141","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2503.15141","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2503.15141","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2503.15141","pdf_url":"https://arxiv.org/pdf/2503.15141","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":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320327336","display_name":"Vlaamse regering","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Object-centric":[0],"representation":[1],"learning":[2],"has":[3],"recently":[4],"been":[5],"successfully":[6],"applied":[7],"to":[8,16,51,67,79,181,199],"real-world":[9,110],"datasets.":[10],"This":[11],"success":[12],"can":[13,48,64],"be":[14,65],"attributed":[15,66],"pretrained":[17,186,209,219],"non-object-centric":[18,206],"foundation":[19],"models,":[20],"whose":[21],"features":[22],"serve":[23],"as":[24,85,124],"reconstruction":[25,86],"targets":[26,31],"for":[27,103,112],"slot":[28,143,159],"attention.":[29],"However,":[30],"must":[32],"remain":[33],"frozen":[34,205],"throughout":[35],"the":[36,44,53,75,113,119,130,147,158,166,179],"training,":[37],"which":[38,63],"sets":[39],"an":[40,125],"upper":[41,148],"bound":[42,149],"on":[43,109,150,187,210],"performance":[45,61,151,197],"object-centric":[46,71,76,105,131,138,202],"models":[47,106,203,220],"attain.":[49],"Attempts":[50],"update":[52],"target":[54,120,167,207],"encoder":[55,78,121,168],"by":[56,96,142,162,169],"bootstrapping":[57],"result":[58],"in":[59,153],"large":[60],"drops,":[62],"its":[68],"lack":[69],"of":[70,129,165,201,212,214],"inductive":[72,139],"biases,":[73],"causing":[74],"model's":[77],"drift":[80],"away":[81],"from":[82,107,190],"representations":[83],"useful":[84],"targets.":[87],"To":[88],"address":[89],"these":[90],"limitations,":[91],"we":[92],"propose":[93],"Object-CEntric":[94],"Pretraining":[95],"Target":[97],"Encoder":[98],"BOotstrapping,":[99],"a":[100,171],"self-distillation":[101],"setup":[102],"training":[104],"scratch,":[108],"data,":[111],"first":[114],"time":[115],"ever.":[116],"In":[117],"OCEBO,":[118],"is":[122],"updated":[123],"exponential":[126],"moving":[127],"average":[128],"model,":[132],"thus":[133],"explicitly":[134],"being":[135],"enriched":[136],"with":[137,204],"biases":[140],"introduced":[141],"attention":[144],"while":[145],"removing":[146],"present":[152],"other":[154],"models.":[155],"We":[156],"mitigate":[157],"collapse":[160],"caused":[161],"random":[163],"initialization":[164],"introducing":[170],"novel":[172],"cross-view":[173],"patch":[174],"filtering":[175],"approach":[176],"that":[177,200],"limits":[178],"supervision":[180],"sufficiently":[182],"informative":[183],"patches.":[184],"When":[185],"241k":[188],"images":[189],"COCO,":[191],"OCEBO":[192],"achieves":[193],"unsupervised":[194],"object":[195],"discovery":[196],"comparable":[198],"encoders":[208],"hundreds":[211],"millions":[213],"images.":[215],"The":[216],"code":[217],"and":[218],"are":[221],"publicly":[222],"available":[223],"at":[224],"https://github.com/djukicn/ocebo.":[225]},"counts_by_year":[],"updated_date":"2026-09-29T14:09:16.559479","created_date":"2025-10-07T00:00:00"}