{"id":"https://openalex.org/W2172130388","doi":"https://doi.org/10.1109/icassp.2012.6289102","title":"Optimization in speech-centric information processing: Criteria and techniques","display_name":"Optimization in speech-centric information processing: Criteria and techniques","publication_year":2012,"publication_date":"2012-03-01","ids":{"openalex":"https://openalex.org/W2172130388","doi":"https://doi.org/10.1109/icassp.2012.6289102","mag":"2172130388"},"language":"en","primary_location":{"id":"doi:10.1109/icassp.2012.6289102","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp.2012.6289102","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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/A5101727205","display_name":"Xiaodong He","orcid":"https://orcid.org/0000-0002-9463-9168"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Xiaodong He","raw_affiliation_strings":["Microsoft Research, Redmond, WA, USA"],"affiliations":[{"raw_affiliation_string":"Microsoft Research, Redmond, WA, USA","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100671324","display_name":"Li Deng","orcid":"https://orcid.org/0000-0002-1014-0790"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Li Deng","raw_affiliation_strings":["Microsoft Research, Redmond, WA, USA"],"affiliations":[{"raw_affiliation_string":"Microsoft Research, Redmond, WA, USA","institution_ids":["https://openalex.org/I1290206253"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5101727205"],"corresponding_institution_ids":["https://openalex.org/I1290206253"],"apc_list":null,"apc_paid":null,"fwci":2.6528,"has_fulltext":false,"cited_by_count":8,"citation_normalized_percentile":{"value":0.90992976,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"5241","last_page":"5244"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9998000264167786,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9998000264167786,"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/T10028","display_name":"Topic Modeling","score":0.9930999875068665,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9930999875068665,"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/computer-science","display_name":"Computer science","score":0.7885525226593018},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5217089056968689},{"id":"https://openalex.org/keywords/speech-processing","display_name":"Speech processing","score":0.5085569620132446},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3512263298034668},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.33929723501205444}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7885525226593018},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5217089056968689},{"id":"https://openalex.org/C61328038","wikidata":"https://www.wikidata.org/wiki/Q3358061","display_name":"Speech processing","level":2,"score":0.5085569620132446},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3512263298034668},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.33929723501205444}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/icassp.2012.6289102","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp.2012.6289102","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},{"id":"pmh:oai:CiteSeerX.psu:10.1.1.306.5697","is_oa":false,"landing_page_url":"http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.306.5697","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"http://research.microsoft.com/pubs/164625/5241.pdf","raw_type":"text"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.699999988079071}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W1506371579","https://openalex.org/W1985554184","https://openalex.org/W2024539680","https://openalex.org/W2050552609","https://openalex.org/W2051669046","https://openalex.org/W2077574412","https://openalex.org/W2098601596","https://openalex.org/W2101105183","https://openalex.org/W2114483840","https://openalex.org/W2136630431","https://openalex.org/W2136727248","https://openalex.org/W2139113820","https://openalex.org/W2144663079","https://openalex.org/W2146574666","https://openalex.org/W2149980590","https://openalex.org/W2151742940","https://openalex.org/W2153653739","https://openalex.org/W2163552996","https://openalex.org/W2168262690","https://openalex.org/W2294759826","https://openalex.org/W2408327647","https://openalex.org/W2917604489","https://openalex.org/W4236796448","https://openalex.org/W4251560691","https://openalex.org/W6656819595","https://openalex.org/W6682235427","https://openalex.org/W6697079997","https://openalex.org/W6697332768","https://openalex.org/W6714065594"],"related_works":["https://openalex.org/W3204019825","https://openalex.org/W2981428355","https://openalex.org/W1834994814","https://openalex.org/W2041273198","https://openalex.org/W1599055764","https://openalex.org/W2131711534","https://openalex.org/W2149163000","https://openalex.org/W2962858469","https://openalex.org/W2289873871","https://openalex.org/W2559040841"],"abstract_inverted_index":{"Automatic":[0],"speech":[1,17,25,133],"recognition":[2],"(ASR)":[3],"is":[4,41,70],"an":[5],"enabling":[6],"technology":[7],"for":[8],"a":[9,68,90,132],"wide":[10],"range":[11],"of":[12,36,52,127,143,157],"information":[13,22,54],"processing":[14,45],"applications":[15],"including":[16,103],"translation,":[18,53],"voice":[19],"search":[20],"(i.e.,":[21],"retrieval":[23],"with":[24,185],"input),":[26],"and":[27,181],"conversational":[28],"understanding.":[29],"In":[30,61,85],"these":[31],"speech-centric":[32],"applications,":[33,63],"the":[34,49,64,75,97,114,119,125,128,137,150,161,173],"output":[35],"ASR":[37,65,104],"as":[38,67],"\u201cnoisy\u201d":[39],"text":[40],"fed":[42],"into":[43],"down-stream":[44,76],"systems":[46],"to":[47,81,118,154],"accomplish":[48],"designated":[50],"tasks":[51],"retrieval,":[55],"or":[56],"natural":[57],"language":[58,139],"understanding,":[59],"etc.":[60],"conventional":[62,162],"model":[66,98],"sub-system":[69,168],"usually":[71],"trained":[72],"without":[73],"considering":[74],"systems.":[77],"This":[78],"often":[79],"leads":[80,153],"sub-optimal":[82],"end-to-end":[83,92,120],"performance.":[84],"this":[86],"paper,":[87],"we":[88],"propose":[89],"unifying":[91],"optimization":[93,175,188],"framework":[94],"in":[95,100,178],"which":[96],"parameters":[99],"all":[101],"sub-systems":[102],"are":[105],"learned":[106],"by":[107],"Extended":[108],"Baum-Welch":[109],"(EBW)":[110],"algorithms":[111,176],"via":[112],"optimizing":[113],"criteria":[115],"directly":[116],"tied":[117],"performance":[121],"measure.":[122],"We":[123,170],"demonstrate":[124],"effectiveness":[126],"proposed":[129,151],"approach":[130],"on":[131,165],"translation":[134,140,158],"task":[135],"using":[136],"spoken":[138],"benchmark":[141],"test":[142],"IWSLT.":[144],"Our":[145],"experimental":[146],"results":[147],"show":[148],"that":[149],"method":[152],"significant":[155],"improvement":[156],"quality":[159],"over":[160],"techniques":[163],"based":[164],"separate":[166],"modular":[167],"design.":[169],"also":[171],"analyze":[172],"EBW-based":[174],"employed":[177],"our":[179],"work":[180],"discuss":[182],"its":[183],"relationship":[184],"other":[186],"popular":[187],"techniques.":[189]},"counts_by_year":[{"year":2017,"cited_by_count":2},{"year":2015,"cited_by_count":2},{"year":2014,"cited_by_count":2},{"year":2012,"cited_by_count":2}],"updated_date":"2026-04-05T17:49:38.594831","created_date":"2025-10-10T00:00:00"}