| Technique | SpokenContent |
|---|---|
| Document | W4743 |
| Name | Phil Garner, Jason Charlesworth |
| philg@cre.canon.co.uk | |
| Type | Application using several basic components |
| External Libraries | None |
| Related Ds/DSs | ProbabilityVector, ProbabilityMatrix |
| Used Ds/DSs | ProbabilityMatrix ProbabilityVector SpokenContentBlock SpokenContentConfusionInfo SpokenContentExtractionInfo SpokenContentHeader SpokenContentIndexEntry SpokenContentLattice SpokenContentLexicon SpokenContentLink SpokenContentNode SpokenContentPhoneIndexEntry SpokenContentPhoneLexicon SpokenContentPhoneLink SpokenContentSpeaker SpokenContentWordIndexEntry SpokenContentWordLexicon SpokenContentWordLink |
| Extraction | NA |
| Client Appl | Search & Retrieval |
| Summary | Demonstrates basic search and retrieval using spoken content. Both data and queries can be word based or phoneme based. |
| Strong Points | Demonstrates how to link an audio or video descriptor with C++ representation to a MDS DS with GenericDS implementation. |
| Limitations | None |
| Known Problems | The MDS elements MediaTime and Individual are ignored by the implementation, even though the GenericDS reads them |
| Parameters | None |