{"id":"https://openalex.org/W7156635889","doi":"https://doi.org/10.48550/arxiv.2604.23314","title":"Learning from Noisy Prompts: Saliency-Guided Prompt Distillation for Robust Segmentation with SAM","display_name":"Learning from Noisy Prompts: Saliency-Guided Prompt Distillation for Robust Segmentation with SAM","publication_year":2026,"publication_date":"2026-04-25","ids":{"openalex":"https://openalex.org/W7156635889","doi":"https://doi.org/10.48550/arxiv.2604.23314"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.23314","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.23314","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.2604.23314","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5037667602","display_name":"Jingxuan Kang","orcid":"https://orcid.org/0000-0003-2605-9265"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Kang, Jingxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134790857","display_name":"Ziqi Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Ziqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089341631","display_name":"Shaoming Zheng","orcid":"https://orcid.org/0000-0001-5628-4311"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Shaoming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134807143","display_name":"Shuang Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Shuang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103076912","display_name":"Uday Patel","orcid":"https://orcid.org/0000-0003-1745-0238"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Patel, Uday Bharat","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134767285","display_name":"Alexander Harry Fitzhugh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fitzhugh, Alexander Harry","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5015532047","display_name":"Phillip Lung","orcid":"https://orcid.org/0000-0002-7459-2323"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lung, Phillip","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134769500","display_name":"Yusuf Kiberu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kiberu, Yusuf","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089562939","display_name":"Nikesh Jathanna","orcid":"https://orcid.org/0000-0002-6114-6068"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jathanna, Nikesh","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134769820","display_name":"Shahnaz Jamil-Copley","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jamil-Copley, Shahnaz","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134764541","display_name":"Bernhard Kainz","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kainz, Bernhard","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5134792784","display_name":"Chen Qin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qin, Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":12,"corresponding_author_ids":["https://openalex.org/A5037667602"],"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/T10036","display_name":"Advanced Neural Network Applications","score":0.2732999920845032,"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":0.2732999920845032,"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/T14510","display_name":"Medical Imaging and Analysis","score":0.23199999332427979,"subfield":{"id":"https://openalex.org/subfields/2204","display_name":"Biomedical Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10052","display_name":"Medical Image Segmentation Techniques","score":0.07800000160932541,"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/prior-probability","display_name":"Prior probability","score":0.7491999864578247},{"id":"https://openalex.org/keywords/segmentation","display_name":"Segmentation","score":0.6043000221252441},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.554099977016449},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.49079999327659607},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.4880000054836273},{"id":"https://openalex.org/keywords/a-priori-and-a-posteriori","display_name":"A priori and a posteriori","score":0.4496000111103058},{"id":"https://openalex.org/keywords/imperfect","display_name":"Imperfect","score":0.4018000066280365},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3917999863624573},{"id":"https://openalex.org/keywords/path","display_name":"Path (computing)","score":0.37959998846054077}],"concepts":[{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.7491999864578247},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7466999888420105},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.7075999975204468},{"id":"https://openalex.org/C89600930","wikidata":"https://www.wikidata.org/wiki/Q1423946","display_name":"Segmentation","level":2,"score":0.6043000221252441},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.554099977016449},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.49079999327659607},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.4880000054836273},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4722999930381775},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.4496000111103058},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.41110000014305115},{"id":"https://openalex.org/C2780310539","wikidata":"https://www.wikidata.org/wiki/Q12547192","display_name":"Imperfect","level":2,"score":0.4018000066280365},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3917999863624573},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.37959998846054077},{"id":"https://openalex.org/C2781181686","wikidata":"https://www.wikidata.org/wiki/Q4226068","display_name":"Coherence (philosophical gambling strategy)","level":2,"score":0.3736000061035156},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.3596000075340271},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.35019999742507935},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.34459999203681946},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.3273000121116638},{"id":"https://openalex.org/C61224824","wikidata":"https://www.wikidata.org/wiki/Q2260434","display_name":"Mixture model","level":2,"score":0.3240000009536743},{"id":"https://openalex.org/C124504099","wikidata":"https://www.wikidata.org/wiki/Q56933","display_name":"Image segmentation","level":3,"score":0.31709998846054077},{"id":"https://openalex.org/C2780966255","wikidata":"https://www.wikidata.org/wiki/Q5474306","display_name":"Foundation (evidence)","level":2,"score":0.3156999945640564},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.3061999976634979},{"id":"https://openalex.org/C31601959","wikidata":"https://www.wikidata.org/wiki/Q931309","display_name":"Medical imaging","level":2,"score":0.30570000410079956},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.30250000953674316},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.28119999170303345},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.2791999876499176},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2574999928474426},{"id":"https://openalex.org/C111370547","wikidata":"https://www.wikidata.org/wiki/Q7451120","display_name":"Sensory cue","level":2,"score":0.2542000114917755},{"id":"https://openalex.org/C16345878","wikidata":"https://www.wikidata.org/wiki/Q107472979","display_name":"Orientation (vector space)","level":2,"score":0.2535000145435333}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.23314","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.23314","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.2604.23314","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.23314","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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Segmentation":[0],"is":[1],"central":[2],"to":[3,101],"clinical":[4,48,190],"diagnosis":[5],"and":[6,42,59,66,115,154,165,174,181],"monitoring,":[7],"yet":[8],"the":[9,21,39,132],"reliability":[10],"of":[11,23,134],"modern":[12],"foundation":[13,186],"models":[14],"in":[15,171,189],"medical":[16],"imaging":[17],"still":[18],"depends":[19],"on":[20,150],"availability":[22],"precise":[24],"prompts.":[25],"The":[26],"Segment":[27],"Anything":[28],"Model":[29],"(SAM)":[30],"offers":[31],"powerful":[32],"zero-shot":[33],"capabilities,":[34],"although":[35],"it":[36],"collapses":[37],"under":[38],"weak,":[40],"generic,":[41],"noisy":[43,117],"prompts":[44,118,195],"that":[45,82,130,158],"dominate":[46],"real":[47],"workflows.":[49],"In":[50],"practice,":[51],"annotations":[52],"such":[53],"as":[54],"centerline":[55],"points":[56],"are":[57,196],"coarse":[58],"ambiguous,":[60],"often":[61],"drifting":[62],"across":[63],"neighboring":[64],"anatomy":[65],"misguiding":[67],"SAM":[68,163],"toward":[69,184],"inconsistent":[70],"or":[71],"incomplete":[72],"masks.":[73],"We":[74],"introduce":[75],"SPD,":[76],"a":[77,97,126,179],"Saliency-Guided":[78],"Prompt":[79,111],"Distillation":[80],"framework":[81],"converts":[83],"these":[84],"unreliable":[85],"cues":[86,120],"into":[87],"robust":[88],"guidance.":[89],"SPD":[90,159,177],"first":[91],"learns":[92],"data-driven":[93],"anatomical":[94,145],"priors":[95,107],"through":[96],"lightweight":[98],"saliency":[99],"head":[100],"obtain":[102],"confident":[103],"localization":[104],"maps.":[105],"These":[106],"then":[108],"drive":[109],"Contextual":[110],"Distillation,":[112],"which":[113],"validates":[114],"enriches":[116],"using":[119],"from":[121],"anatomically":[122],"adjacent":[123],"slices,":[124],"producing":[125],"consensus":[127],"prompt":[128],"set":[129],"matches":[131],"behavior":[133],"expert":[135],"reasoning.":[136],"A":[137],"Pairwise":[138],"Slice":[139],"Consistency":[140],"objective":[141],"further":[142],"enforces":[143],"local":[144],"coherence":[146],"during":[147],"segmentation.":[148],"Experiments":[149],"four":[151],"challenging":[152],"MRI":[153],"CT":[155],"benchmarks":[156],"demonstrate":[157],"consistently":[160],"outperforms":[161],"existing":[162],"adaptations":[164],"supervised":[166],"baselines,":[167],"delivering":[168],"large":[169],"gains":[170],"both":[172],"region-based":[173],"boundary-based":[175],"metrics.":[176],"provides":[178],"practical":[180],"principled":[182],"path":[183],"reliable":[185],"model":[187],"deployment":[188],"environments":[191],"where":[192],"only":[193],"imperfect":[194],"available.":[197]},"counts_by_year":[],"updated_date":"2026-04-29T06:16:36.941037","created_date":"2026-04-29T00:00:00"}