Voxel-based Viterbi Active Speaker Tracking (V-VAST) with best view selection for video lecture post-production
An automated system is presented for reducing a multi-view lecture recording into a single view video containing a best view summary of active speakers. The system uses skin color detection and voxel-based analysis in locating likely speaker locations. Using time-delay estimates from multiple micro...
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Tagungsbericht |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 2299 |
|---|---|
| container_issue | |
| container_start_page | 2296 |
| container_title | |
| container_volume | |
| creator | Kelly, Damien Kokaram, Anil Boland, Frank |
| description | An automated system is presented for reducing a multi-view lecture recording into a single view video containing a best view summary of active speakers. The system uses skin color detection and voxel-based analysis in locating likely speaker locations. Using time-delay estimates from multiple micro phones, speech activity is analyzed for each speaker position. The Viterbi algorithm is then used to estimate a track of the active speaker which maximizes the observed speech activity. This novel approach is termed Voxel-based Viterbi Active Speaker Tracking (V-VAST) and is shown to track speakers with an accuracy of 0.23m. Using the tracking information, the system then extracts from the available camera views the most frontal face view of the active speaker to display. |
| doi_str_mv | 10.1109/ICASSP.2011.5946941 |
| format | Conference Proceeding |
| fullrecord | <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_5946941</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>5946941</ieee_id><sourcerecordid>5946941</sourcerecordid><originalsourceid>FETCH-LOGICAL-i175t-f1cd9a013fc0783be22e12c465aa9f439f0444aaef6fc608b3b31f2568846cea3</originalsourceid><addsrcrecordid>eNo1UMlOwzAUNJtEKf2CXnyEg4t3x8cKsUmVQEqJuFWO8wympYmctIW_J0CZy0izPOkNQmNGJ4xRe_VwPc3zpwmnjE2UldpKdoDOmFTGUCWsOUQDLowlzNKXIzSyJvv3MnqMBkxxSjST9hSN2vad9tDcGGUHqC7qT1iR0rVQ4SJ2kMqIp76LW8B5A24JCc-T88u4fsUXBSmm-fwS72L3hktoO7yNsMMtrKCv1Gsc6tRLFdT4R9kkwE3ddqRJdbX5TZyjk-BWLYz2PETPtzfz63sye7zrv5yRyIzqSGC-so4yETw1mSiBc2DcS62cs0EKG6iU0jkIOnhNs1KUggWudJZJ7cGJIRr_3Y0AsGhS_HDpa7HfTnwD1Itf7Q</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Voxel-based Viterbi Active Speaker Tracking (V-VAST) with best view selection for video lecture post-production</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Kelly, Damien ; Kokaram, Anil ; Boland, Frank</creator><creatorcontrib>Kelly, Damien ; Kokaram, Anil ; Boland, Frank</creatorcontrib><description>An automated system is presented for reducing a multi-view lecture recording into a single view video containing a best view summary of active speakers. The system uses skin color detection and voxel-based analysis in locating likely speaker locations. Using time-delay estimates from multiple micro phones, speech activity is analyzed for each speaker position. The Viterbi algorithm is then used to estimate a track of the active speaker which maximizes the observed speech activity. This novel approach is termed Voxel-based Viterbi Active Speaker Tracking (V-VAST) and is shown to track speakers with an accuracy of 0.23m. Using the tracking information, the system then extracts from the available camera views the most frontal face view of the active speaker to display.</description><identifier>ISSN: 1520-6149</identifier><identifier>ISBN: 9781457705380</identifier><identifier>ISBN: 1457705389</identifier><identifier>EISSN: 2379-190X</identifier><identifier>EISBN: 1457705397</identifier><identifier>EISBN: 9781457705373</identifier><identifier>EISBN: 9781457705397</identifier><identifier>EISBN: 1457705370</identifier><identifier>DOI: 10.1109/ICASSP.2011.5946941</identifier><language>eng</language><publisher>IEEE</publisher><subject>Audio-Visual Tracking ; Cameras ; Face ; Microphones ; Multi-camera ; Multi-microphone ; Skin ; Speech ; Three dimensional displays ; Time-Delay Estimation ; Viterbi ; Viterbi algorithm</subject><ispartof>2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2011, p.2296-2299</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/5946941$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,776,780,785,786,2052,27902,54895</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/5946941$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Kelly, Damien</creatorcontrib><creatorcontrib>Kokaram, Anil</creatorcontrib><creatorcontrib>Boland, Frank</creatorcontrib><title>Voxel-based Viterbi Active Speaker Tracking (V-VAST) with best view selection for video lecture post-production</title><title>2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)</title><addtitle>ICASSP</addtitle><description>An automated system is presented for reducing a multi-view lecture recording into a single view video containing a best view summary of active speakers. The system uses skin color detection and voxel-based analysis in locating likely speaker locations. Using time-delay estimates from multiple micro phones, speech activity is analyzed for each speaker position. The Viterbi algorithm is then used to estimate a track of the active speaker which maximizes the observed speech activity. This novel approach is termed Voxel-based Viterbi Active Speaker Tracking (V-VAST) and is shown to track speakers with an accuracy of 0.23m. Using the tracking information, the system then extracts from the available camera views the most frontal face view of the active speaker to display.</description><subject>Audio-Visual Tracking</subject><subject>Cameras</subject><subject>Face</subject><subject>Microphones</subject><subject>Multi-camera</subject><subject>Multi-microphone</subject><subject>Skin</subject><subject>Speech</subject><subject>Three dimensional displays</subject><subject>Time-Delay Estimation</subject><subject>Viterbi</subject><subject>Viterbi algorithm</subject><issn>1520-6149</issn><issn>2379-190X</issn><isbn>9781457705380</isbn><isbn>1457705389</isbn><isbn>1457705397</isbn><isbn>9781457705373</isbn><isbn>9781457705397</isbn><isbn>1457705370</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2011</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNo1UMlOwzAUNJtEKf2CXnyEg4t3x8cKsUmVQEqJuFWO8wympYmctIW_J0CZy0izPOkNQmNGJ4xRe_VwPc3zpwmnjE2UldpKdoDOmFTGUCWsOUQDLowlzNKXIzSyJvv3MnqMBkxxSjST9hSN2vad9tDcGGUHqC7qT1iR0rVQ4SJ2kMqIp76LW8B5A24JCc-T88u4fsUXBSmm-fwS72L3hktoO7yNsMMtrKCv1Gsc6tRLFdT4R9kkwE3ddqRJdbX5TZyjk-BWLYz2PETPtzfz63sye7zrv5yRyIzqSGC-so4yETw1mSiBc2DcS62cs0EKG6iU0jkIOnhNs1KUggWudJZJ7cGJIRr_3Y0AsGhS_HDpa7HfTnwD1Itf7Q</recordid><startdate>201105</startdate><enddate>201105</enddate><creator>Kelly, Damien</creator><creator>Kokaram, Anil</creator><creator>Boland, Frank</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>201105</creationdate><title>Voxel-based Viterbi Active Speaker Tracking (V-VAST) with best view selection for video lecture post-production</title><author>Kelly, Damien ; Kokaram, Anil ; Boland, Frank</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i175t-f1cd9a013fc0783be22e12c465aa9f439f0444aaef6fc608b3b31f2568846cea3</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2011</creationdate><topic>Audio-Visual Tracking</topic><topic>Cameras</topic><topic>Face</topic><topic>Microphones</topic><topic>Multi-camera</topic><topic>Multi-microphone</topic><topic>Skin</topic><topic>Speech</topic><topic>Three dimensional displays</topic><topic>Time-Delay Estimation</topic><topic>Viterbi</topic><topic>Viterbi algorithm</topic><toplevel>online_resources</toplevel><creatorcontrib>Kelly, Damien</creatorcontrib><creatorcontrib>Kokaram, Anil</creatorcontrib><creatorcontrib>Boland, Frank</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Kelly, Damien</au><au>Kokaram, Anil</au><au>Boland, Frank</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Voxel-based Viterbi Active Speaker Tracking (V-VAST) with best view selection for video lecture post-production</atitle><btitle>2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)</btitle><stitle>ICASSP</stitle><date>2011-05</date><risdate>2011</risdate><spage>2296</spage><epage>2299</epage><pages>2296-2299</pages><issn>1520-6149</issn><eissn>2379-190X</eissn><isbn>9781457705380</isbn><isbn>1457705389</isbn><eisbn>1457705397</eisbn><eisbn>9781457705373</eisbn><eisbn>9781457705397</eisbn><eisbn>1457705370</eisbn><abstract>An automated system is presented for reducing a multi-view lecture recording into a single view video containing a best view summary of active speakers. The system uses skin color detection and voxel-based analysis in locating likely speaker locations. Using time-delay estimates from multiple micro phones, speech activity is analyzed for each speaker position. The Viterbi algorithm is then used to estimate a track of the active speaker which maximizes the observed speech activity. This novel approach is termed Voxel-based Viterbi Active Speaker Tracking (V-VAST) and is shown to track speakers with an accuracy of 0.23m. Using the tracking information, the system then extracts from the available camera views the most frontal face view of the active speaker to display.</abstract><pub>IEEE</pub><doi>10.1109/ICASSP.2011.5946941</doi><tpages>4</tpages></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 1520-6149 |
| ispartof | 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2011, p.2296-2299 |
| issn | 1520-6149 2379-190X |
| language | eng |
| recordid | cdi_ieee_primary_5946941 |
| source | IEEE Electronic Library (IEL) Conference Proceedings |
| subjects | Audio-Visual Tracking Cameras Face Microphones Multi-camera Multi-microphone Skin Speech Three dimensional displays Time-Delay Estimation Viterbi Viterbi algorithm |
| title | Voxel-based Viterbi Active Speaker Tracking (V-VAST) with best view selection for video lecture post-production |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-10T05%3A44%3A12IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Voxel-based%20Viterbi%20Active%20Speaker%20Tracking%20(V-VAST)%20with%20best%20view%20selection%20for%20video%20lecture%20post-production&rft.btitle=2011%20IEEE%20International%20Conference%20on%20Acoustics,%20Speech%20and%20Signal%20Processing%20(ICASSP)&rft.au=Kelly,%20Damien&rft.date=2011-05&rft.spage=2296&rft.epage=2299&rft.pages=2296-2299&rft.issn=1520-6149&rft.eissn=2379-190X&rft.isbn=9781457705380&rft.isbn_list=1457705389&rft_id=info:doi/10.1109/ICASSP.2011.5946941&rft_dat=%3Cieee_6IE%3E5946941%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&rft.eisbn=1457705397&rft.eisbn_list=9781457705373&rft.eisbn_list=9781457705397&rft.eisbn_list=1457705370&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=5946941&rfr_iscdi=true |