{"id":"https://openalex.org/W7166810890","doi":"https://doi.org/10.18653/v1/2026.findings-acl.538","title":"Waking Up Blind: Cold-Start Optimization of Supervision-Free Agentic Trajectories for Grounded Visual Perception","display_name":"Waking Up Blind: Cold-Start Optimization of Supervision-Free Agentic Trajectories for Grounded Visual Perception","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166810890","doi":"https://doi.org/10.18653/v1/2026.findings-acl.538"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.538","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.538","pdf_url":"https://aclanthology.org/2026.findings-acl.538.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.538.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5124005971","display_name":"Ashutosh Bajpai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ashutosh Bajpai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139771597","display_name":"Tamal Majumder","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tamal Majumder","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139781329","display_name":"Akshay Nambi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Akshay Nambi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139822722","display_name":"Tanmoy Chakraborty","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tanmoy Chakraborty","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.86003303,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"11070","last_page":"11093"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.48500001430511475,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.48500001430511475,"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":0.18629999458789825,"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":0.04190000146627426,"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/perception","display_name":"Perception","score":0.5022000074386597},{"id":"https://openalex.org/keywords/visual-perception","display_name":"Visual perception","score":0.3806000053882599},{"id":"https://openalex.org/keywords/eye-tracking","display_name":"Eye tracking","score":0.3547999858856201},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.3425999879837036},{"id":"https://openalex.org/keywords/cognition","display_name":"Cognition","score":0.3077000081539154},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.2721000015735626}],"concepts":[{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.6335999965667725},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.5479999780654907},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.5022000074386597},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.38609999418258667},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.3806000053882599},{"id":"https://openalex.org/C56461940","wikidata":"https://www.wikidata.org/wiki/Q970687","display_name":"Eye tracking","level":2,"score":0.3547999858856201},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.3425999879837036},{"id":"https://openalex.org/C169900460","wikidata":"https://www.wikidata.org/wiki/Q2200417","display_name":"Cognition","level":2,"score":0.3077000081539154},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.2897000014781952},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2721000015735626},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.26820001006126404},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.2662999927997589},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2660999894142151},{"id":"https://openalex.org/C156325361","wikidata":"https://www.wikidata.org/wiki/Q1152864","display_name":"Grounded theory","level":3,"score":0.25839999318122864},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.257099986076355},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.2540000081062317}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.538","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.538","pdf_url":"https://aclanthology.org/2026.findings-acl.538.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.538","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.538","pdf_url":"https://aclanthology.org/2026.findings-acl.538.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320308943","display_name":"Microsoft Research","ror":"https://ror.org/00d0nc645"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166810890.pdf","grobid_xml":"https://content.openalex.org/works/W7166810890.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Small":[0],"Vision-Language":[1],"Models":[2],"(SVLMs)":[3],"are":[4],"efficient":[5,149],"task":[6,86,136],"controllers":[7],"but":[8],"often":[9],"suffer":[10],"from":[11,155],"visual":[12,76],"brittleness":[13],"and":[14,90,125,142],"poor":[15],"tool":[16,91,114,143],"orchestration.They":[17],"typically":[18],"require":[19],"expensive":[20],"supervised":[21],"trajectory":[22],"tuning":[23],"to":[24,65,95,112,140],"mitigate":[25],"these":[26],"deficits.In":[27],"this":[28],"work,":[29],"we":[30],"propose":[31],"Self-supervised":[32],"Perception":[33],"Enabled":[34],"by":[35,138,145],"Cascaded":[36],"Tool":[37,105],"Rollout":[38],"Alignment":[39],"(SPECTRA),":[40],"a":[41,59,79,109],"supervision-free":[42],"framework":[43],"that":[44,62,83,130,152],"bootstraps":[45],"agentic":[46,133],"capabilities":[47],"via":[48],"Coldstart":[49],"Reinforcement":[50],"Learning":[51],"for":[52],"SVLMs.SPECTRA":[53],"enforces":[54],"Soft":[55],"Structured":[56],"Multi-turn":[57],"Rollouts,":[58],"topological":[60],"constraint":[61],"directs":[63],"agents":[64,151],"explicitly":[66],"sequence":[67],"toolderived":[68],"evidence":[69],"before":[70],"synthesis,":[71],"effectively":[72,154],"grounding":[73],"reasoning":[74],"in":[75,116],"observations.We":[77],"employ":[78],"multi-objective":[80],"reward":[81],"signal":[82],"simultaneously":[84],"maximizes":[85],"correctness,":[87],"rollout":[88],"structure,":[89],"utility,":[92],"enabling":[93,147],"agent":[94],"self-discover":[96],"robust":[97],"behaviors":[98],"without":[99],"human":[100],"preference":[101],"labels.We":[102],"further":[103],"introduce":[104],"Instrumental":[106],"Utility":[107],"(TIU),":[108],"novel":[110],"metric":[111],"quantify":[113],"efficacy":[115],"the":[117],"absence":[118],"of":[119],"ground":[120],"truth.Extensive":[121],"evaluations":[122],"across":[123],"composite":[124],"out-of-distribution":[126],"(MMMU-Pro)":[127],"benchmarks":[128],"demonstrate":[129],"SPECTRA":[131],"boosts":[132],"trajectories,":[134],"improving":[135],"accuracy":[137],"up":[139],"5%":[141],"efficiency":[144],"9%,":[146],"more":[148],"multimodal":[150],"learn":[153],"environmental":[156],"interaction":[157],"alone.":[158]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}