en-de
en-es
en-fr
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
The Rich Transcription 2005 Spring Meeting Recognition Evaluation
Published on Feb 4, 20252964 Views
Related categories
Presentation
Welcome to the Rich Transcription 2005 Spring Meeting Recognition Evaluation Workshop00:02
Today’s Agenda00:18
Administrative Points00:27
The Rich Transcription 2005 Spring Meeting Recognition Evaluation00:36
Overview00:50
The Rich Transcription Task01:05
Rich Transcription Evaluation Series01:41
NIST STT Benchmark Test History – May. ’0502:17
Research Opportunities in the Meeting Domain03:42
RT-05S Evaluation Tasks04:18
Five System Input Conditions04:40
Training/Development Corpora07:04
RT-05S Evaluation Test Corpora: Conference Room Test Set07:57
RT-05S Evaluation Test Corpora: Lecture Room Test Set10:18
RT-05S Evaluation Participants12:26
Diarization “Who Spoke When” (SPKR) Task12:47
SPKR System Evaluation Method13:51
RT-05S SPKR Results Primary Systems, Non-Overlapping Speech15:51
Lecture Room Results: Broken Down by Excerpt Type17:11
Historical Best System SPKR Performance on Conference Data18:38
Diarization “Speech Activity Detection” (SAD) Task19:16
SAD System Evaluation Method20:56
RT-05S SAD Results Primary Systems21:22
Speech-To-Text (STT) Task22:15
STT System Evaluation Method22:48
Simultaneous Speech for STT23:47
RT-05S STT Results Primary Systems (Incl. overlaps)24:52
Historical STT Performance in the Meeting Domain26:19
STT Error Rates: Effect of Simultaneous Speech 27:01
Diarization “Source Localization” (SLOC) Task27:39
SLOC System Evaluation Method28:47
R-05S SLOC Results Primary Systems29:44
Summary31:49
Issues for RT-06 Meeting Eval32:26
Proposal for RT-0632:56