AT405924T

Robust and invariant audio pattern matching

Abstract

The invention claims an optical with quickly and welded for accurately two audio samples matching, and whether to avoid for playing variation of the invention technologies of the transformations. The audio sample between the relation of first matched to the output different specimen a fingerprint things. (210) For generating a set of (230) fingerprint things to each audio sample (231, 232) of each other in occurred a special position (242) is. The position (242) decisions panel according to the audio sample (210) contents, wherein the fingerprint things (232) provided with a or closing the special position (242) or more partial for (222). Then to each time to matching fingerprint object decision relative value. And a relative value histogram. Discovering and counting a obvious peak, and two audio samples matched with the inlet of.

Term

No projected expiry on record.

  1. Priority
  2. Filed
  3. Granted
  4. Today

26 members in 16 offices

This recordMember, by publication datePriority claim6 more offices are in the list below

Priority claims1

Priority claims
DocumentOfficeKindDate
37605502United States of AmericaP

Members26

Family members, oldest first
DocumentOfficeKind
CA2483104A1CanadaA1
WO03091990A1World Intellectual Property Organization (WIPO)A1
AU2003230993A1AustraliaA1
TW200307205ATaiwan Province of ChinaA
KR20050010763ARepublic of KoreaA
BR0309598ABrazilA
EP1504445A1European Patent Office (EPO)A1
CN1647160AChinaA
JP2005524108AJapanA
US2005177372A1United States of AmericaA1
EP1504445A4European Patent Office (EPO)A4
HK1073382A1Hong Kong, ChinaA1
TWI269196BTaiwan Province of ChinaB
CN1315110CChinaC
KR100820385B1Republic of KoreaB1
EP1504445B1European Patent Office (EPO)B1
AT405924TThis recordAustriaT
ATE405924T1AustriaT1
DE60323086D1GermanyD1
PT1504445EPortugalE
DK1504445T3DenmarkT3
ES2312772T3SpainT3
US2009265174A9United States of AmericaA9
US7627477B2United States of AmericaB2
JP4425126B2JapanB2
CA2483104CCanadaC