{"id":"https://openalex.org/W7151522617","doi":"https://doi.org/10.1109/icmla66185.2025.00210","title":"Pi-talk: Edge-Only, Adapter-Tuned Multimodal Small Language Model for Safe, Real-Time In-Vehicle Dialogue","display_name":"Pi-talk: Edge-Only, Adapter-Tuned Multimodal Small Language Model for Safe, Real-Time In-Vehicle Dialogue","publication_year":2025,"publication_date":"2025-12-03","ids":{"openalex":"https://openalex.org/W7151522617","doi":"https://doi.org/10.1109/icmla66185.2025.00210"},"language":null,"primary_location":{"id":"doi:10.1109/icmla66185.2025.00210","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icmla66185.2025.00210","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 International Conference on Machine Learning and Applications (ICMLA)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5005124763","display_name":"Alex Pissinou Makki","orcid":null},"institutions":[{"id":"https://openalex.org/I170897317","display_name":"Duke University","ror":"https://ror.org/00py81415","country_code":"US","type":"education","lineage":["https://openalex.org/I170897317"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Alex Pissinou Makki","raw_affiliation_strings":["Duke University,Department of Electrical and Computer Engineering,Durham,NC,USA"],"affiliations":[{"raw_affiliation_string":"Duke University,Department of Electrical and Computer Engineering,Durham,NC,USA","institution_ids":["https://openalex.org/I170897317"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020766546","display_name":"Vahid Tarokh","orcid":"https://orcid.org/0000-0003-2994-6302"},"institutions":[{"id":"https://openalex.org/I170897317","display_name":"Duke University","ror":"https://ror.org/00py81415","country_code":"US","type":"education","lineage":["https://openalex.org/I170897317"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Vahid Tarokh","raw_affiliation_strings":["Duke University,Department of Electrical and Computer Engineering,Durham,NC,USA"],"affiliations":[{"raw_affiliation_string":"Duke University,Department of Electrical and Computer Engineering,Durham,NC,USA","institution_ids":["https://openalex.org/I170897317"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133124601","display_name":"Louis Lago Enamorado","orcid":null},"institutions":[{"id":"https://openalex.org/I19700959","display_name":"Florida International University","ror":"https://ror.org/02gz6gg07","country_code":"US","type":"education","lineage":["https://openalex.org/I19700959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Louis Lago Enamorado","raw_affiliation_strings":["Florida International University,Knight Foundation School of Computing and Information Sciences,Miami,FL,USA"],"affiliations":[{"raw_affiliation_string":"Florida International University,Knight Foundation School of Computing and Information Sciences,Miami,FL,USA","institution_ids":["https://openalex.org/I19700959"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133070426","display_name":"Carlos Carroz","orcid":null},"institutions":[{"id":"https://openalex.org/I19700959","display_name":"Florida International University","ror":"https://ror.org/02gz6gg07","country_code":"US","type":"education","lineage":["https://openalex.org/I19700959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Carlos Carroz","raw_affiliation_strings":["Florida International University,Knight Foundation School of Computing and Information Sciences,Miami,FL,USA"],"affiliations":[{"raw_affiliation_string":"Florida International University,Knight Foundation School of Computing and Information Sciences,Miami,FL,USA","institution_ids":["https://openalex.org/I19700959"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043735527","display_name":"M. Dolores Mart\u00ednez-Mena Garc\u00eda","orcid":null},"institutions":[{"id":"https://openalex.org/I2802326326","display_name":"The University of Texas Rio Grande Valley","ror":"https://ror.org/02p5xjf12","country_code":"US","type":"education","lineage":["https://openalex.org/I2802326326"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Miguel Garcia","raw_affiliation_strings":["The University of Texas Rio Grande Valley,Department of Computer Science,Edinburg,TX,USA"],"affiliations":[{"raw_affiliation_string":"The University of Texas Rio Grande Valley,Department of Computer Science,Edinburg,TX,USA","institution_ids":["https://openalex.org/I2802326326"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5133064707","display_name":"Niki Pissinou","orcid":null},"institutions":[{"id":"https://openalex.org/I19700959","display_name":"Florida International University","ror":"https://ror.org/02gz6gg07","country_code":"US","type":"education","lineage":["https://openalex.org/I19700959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Niki Pissinou","raw_affiliation_strings":["Florida International University,Knight Foundation School of Computing and Information Sciences,Miami,FL,USA"],"affiliations":[{"raw_affiliation_string":"Florida International University,Knight Foundation School of Computing and Information Sciences,Miami,FL,USA","institution_ids":["https://openalex.org/I19700959"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5005124763"],"corresponding_institution_ids":["https://openalex.org/I170897317"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.87290716,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1376","last_page":"1383"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12031","display_name":"Speech and dialogue systems","score":0.8306999802589417,"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/T12031","display_name":"Speech and dialogue systems","score":0.8306999802589417,"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/T10525","display_name":"Human-Automation Interaction and Safety","score":0.027699999511241913,"subfield":{"id":"https://openalex.org/subfields/3207","display_name":"Social Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.025599999353289604,"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/language-model","display_name":"Language model","score":0.33340001106262207},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.31869998574256897},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.2989000082015991},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.2694999873638153},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.2596000134944916}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6430000066757202},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.43939998745918274},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.43070000410079956},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.42899999022483826},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.33340001106262207},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.31869998574256897},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2989000082015991},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2596000134944916},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.2515999972820282}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icmla66185.2025.00210","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icmla66185.2025.00210","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 International Conference on Machine Learning and Applications (ICMLA)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":7,"referenced_works":["https://openalex.org/W2045128528","https://openalex.org/W2568772110","https://openalex.org/W4236853429","https://openalex.org/W4385245566","https://openalex.org/W4394994587","https://openalex.org/W4410825156","https://openalex.org/W4410887653"],"related_works":[],"abstract_inverted_index":{"Natural-language":[0],"interaction":[1],"between":[2],"passengers":[3],"and":[4,12,31,64,93,110,115,126,132,160,163,183,186,211],"autonomous":[5,193],"vehicles":[6],"is":[7,25,100,199],"essential":[8],"for":[9,79,192],"trust,":[10],"safety,":[11],"user":[13],"experience,":[14],"but":[15],"deploying":[16],"Large":[17],"Language":[18,47],"Models":[19],"(LLMs)":[20],"on":[21,52,147,218],"automotive":[22],"edge":[23],"platforms":[24],"constrained":[26],"by":[27],"compute,":[28],"memory,":[29],"energy,":[30],"privacy.":[32],"We":[33,118,150],"present":[34],"Pi-talk,":[35],"an":[36,133,224],"edge-only":[37,177],"system":[38],"that":[39,72,107,173,208],"enables":[40],"real-time":[41,104],"passenger\u2013vehicle":[42,205],"dialogue":[43,206],"using":[44,112],"a":[45,68,80,103,213,228],"Small":[46],"Model":[48],"(SLM)":[49],"running":[50],"entirely":[51,217],"embedded":[53],"hardware.":[54,149],"Pi-talk":[55,174,198],"performs":[56],"multimodal":[57,204],"fusion":[58],"of":[59,88],"onboard":[60],"camera,":[61],"ultrasonic":[62],"distance,":[63],"navigation":[65],"context":[66],"via":[67],"lightweight":[69],"encoder\u2013adapter":[70],"module":[71],"aligns":[73],"modalities":[74],"into":[75],"compact":[76],"semantic":[77],"tokens":[78],"pre-trained":[81],"SLM.":[82],"The":[83],"SLM":[84,122],"produces":[85],"context-aware":[86],"explanations":[87],"driving":[89],"decisions,":[90],"route":[91],"options,":[92],"situational":[94],"updates":[95],"without":[96],"cloud":[97],"connectivity.":[98],"Safety":[99],"enforced":[101],"through":[102],"safety":[105,189,230],"envelope":[106,190],"gates":[108],"responses":[109],"actions":[111],"distance":[113],"thresholds":[114],"timing":[116],"constraints.":[117],"further":[119],"adapter-tune":[120],"the":[121,188,201],"(on-device":[123],"or":[124],"offline)":[125],"deploy":[127],"it":[128],"with":[129,223],"INT8":[130],"quantization":[131],"Open":[134],"Neural":[135],"Network":[136],"Exchange":[137],"(ONNX)":[138],"runtime":[139,229],"to":[140],"achieve":[141],"efficient":[142],"batch":[143],"=":[144],"1":[145],"inference":[146,178],"Raspberry-Pi\u2013class":[148],"evaluate":[151],"task":[152],"quality":[153],"(evaluation":[154],"loss),":[155],"end-to-end":[156],"latency,":[157],"CPU":[158],"utilization,":[159],"memory":[161],"footprint,":[162],"include":[164],"ablations":[165],"contrasting":[166],"unimodal":[167],"vs.":[168],"fused":[169],"inputs.":[170],"Results":[171],"show":[172],"sustains":[175],"few-second,":[176],"while":[179,226],"meeting":[180],"stringent":[181],"resource":[182],"latency":[184],"limits":[185],"maintaining":[187],"required":[191],"operation.":[194],"To":[195],"our":[196],"knowledge,":[197],"among":[200],"first":[202],"edge-only,":[203],"systems":[207],"both":[209],"fine-tune":[210],"run":[212],"small":[214],"language":[215],"model":[216],"Raspberry":[219],"Pi\u2013class,":[220],"CPU-only":[221],"hardware":[222],"explicit":[225],"enforcing":[227],"envelope.":[231]},"counts_by_year":[],"updated_date":"2026-04-09T06:08:40.794217","created_date":"2026-04-08T00:00:00"}