{"id":"https://openalex.org/W4403791371","doi":"https://doi.org/10.1145/3664647.3681690","title":"Dissecting Temporal Understanding in Text-to-Audio Retrieval","display_name":"Dissecting Temporal Understanding in Text-to-Audio Retrieval","publication_year":2024,"publication_date":"2024-10-26","ids":{"openalex":"https://openalex.org/W4403791371","doi":"https://doi.org/10.1145/3664647.3681690"},"language":"en","primary_location":{"id":"doi:10.1145/3664647.3681690","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3664647.3681690","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 32nd ACM International Conference on Multimedia","raw_type":"proceedings-article"},"type":"conference-paper","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/A5036345396","display_name":"Andreea-Maria Oncescu","orcid":"https://orcid.org/0000-0001-9730-3873"},"institutions":[{"id":"https://openalex.org/I2802123492","display_name":"Oxford Research Group","ror":"https://ror.org/00z4w4f29","country_code":"GB","type":"nonprofit","lineage":["https://openalex.org/I2802123492"]},{"id":"https://openalex.org/I40120149","display_name":"University of Oxford","ror":"https://ror.org/052gg0110","country_code":"GB","type":"education","lineage":["https://openalex.org/I40120149"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Andreea-Maria Oncescu","raw_affiliation_strings":["Visual Geometry Group, University of Oxford, Oxford, United Kingdom"],"raw_orcid":"https://orcid.org/0000-0001-9730-3873","affiliations":[{"raw_affiliation_string":"Visual Geometry Group, University of Oxford, Oxford, United Kingdom","institution_ids":["https://openalex.org/I2802123492","https://openalex.org/I40120149"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109730096","display_name":"Jo\u00e3o F. Henriques","orcid":"https://orcid.org/0000-0002-2478-2102"},"institutions":[{"id":"https://openalex.org/I2802123492","display_name":"Oxford Research Group","ror":"https://ror.org/00z4w4f29","country_code":"GB","type":"nonprofit","lineage":["https://openalex.org/I2802123492"]},{"id":"https://openalex.org/I40120149","display_name":"University of Oxford","ror":"https://ror.org/052gg0110","country_code":"GB","type":"education","lineage":["https://openalex.org/I40120149"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Jo\u00e3o F. Henriques","raw_affiliation_strings":["Visual Geometry Group, University of Oxford, Oxford, United Kingdom"],"raw_orcid":"https://orcid.org/0000-0002-2478-2102","affiliations":[{"raw_affiliation_string":"Visual Geometry Group, University of Oxford, Oxford, United Kingdom","institution_ids":["https://openalex.org/I2802123492","https://openalex.org/I40120149"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001366802","display_name":"A. Sophia Koepke","orcid":"https://orcid.org/0000-0002-5807-0576"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"A. Sophia Koepke","raw_affiliation_strings":["T\u00fcbingen AI Center, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0000-0002-5807-0576","affiliations":[{"raw_affiliation_string":"T\u00fcbingen AI Center, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.8343,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.7288898,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"9535","last_page":"9543"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11309","display_name":"Music and Audio Processing","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T11309","display_name":"Music and Audio Processing","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T11439","display_name":"Video Analysis and Summarization","score":0.9896000027656555,"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/T11349","display_name":"Music Technology and Sound Studies","score":0.9821000099182129,"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/computer-science","display_name":"Computer science","score":0.7840166091918945},{"id":"https://openalex.org/keywords/information-retrieval","display_name":"Information retrieval","score":0.447131872177124},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.3839293122291565},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.37568747997283936},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.33691245317459106}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7840166091918945},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.447131872177124},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3839293122291565},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.37568747997283936},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33691245317459106}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3664647.3681690","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3664647.3681690","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 32nd ACM International Conference on Multimedia","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.5600000023841858,"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":16,"referenced_works":["https://openalex.org/W398859631","https://openalex.org/W2074188409","https://openalex.org/W2560662850","https://openalex.org/W2940092410","https://openalex.org/W3015591594","https://openalex.org/W3163547718","https://openalex.org/W4221157007","https://openalex.org/W4367359628","https://openalex.org/W4372260330","https://openalex.org/W4372266552","https://openalex.org/W4375869156","https://openalex.org/W4385488967","https://openalex.org/W4392909554","https://openalex.org/W4392910827","https://openalex.org/W4402702962","https://openalex.org/W4402727693"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Recent":[0],"advancements":[1],"in":[2,63],"machine":[3],"learning":[4],"have":[5],"fueled":[6],"research":[7],"on":[8,84,121],"multimodal":[9],"tasks,":[10],"such":[11],"as":[12],"for":[13,81,102],"instance":[14],"text-to-video":[15],"and":[16,30,35,45,87,128],"text-to-audio":[17,67,82],"retrieval.":[18,68],"These":[19],"tasks":[20],"require":[21],"models":[22,38,118],"to":[23,41,119],"understand":[24],"the":[25,53,64,73,85,122],"semantic":[26],"content":[27],"of":[28,56,66,77,106,125],"video":[29],"audio":[31],"data,":[32],"including":[33],"objects,":[34],"characters.":[36],"The":[37],"also":[39],"need":[40],"learn":[42],"spatial":[43],"arrangements":[44],"temporal":[46,54,74,104,123],"relationships.":[47],"In":[48,69],"this":[49],"work,":[50],"we":[51,71,91,110],"analyse":[52],"ordering":[55,124],"sounds,":[57],"which":[58],"is":[59],"an":[60],"understudied":[61],"problem":[62],"context":[65],"particular,":[70],"dissect":[72],"understanding":[75],"capabilities":[76,105],"a":[78,93,99,112],"state-of-the-art":[79],"model":[80],"retrieval":[83],"AudioCaps":[86],"Clotho":[88],"datasets.":[89],"Additionally,":[90],"introduce":[92],"synthetic":[94],"text-audio":[95,117],"dataset":[96],"that":[97,115],"provides":[98],"controlled":[100],"setting":[101],"evaluating":[103],"recent":[107],"models.":[108],"Lastly,":[109],"present":[111],"loss":[113],"function":[114],"encourages":[116],"focus":[120],"events.":[126],"Code":[127],"data":[129],"are":[130],"available":[131],"at":[132],"https://github.com/oncescuandreea/DTU_text_audio.":[133]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}