Method and device for pause limit values in speech recognition

• US 7,366,667 B2
• Filed: 12/21/2001
• Issued: 04/29/2008
• Est. Priority Date: 12/21/2001
• Status: Expired due to Term

• Current Assignee: Telefonaktiebolaget LM Ericsson
• Original Assignee: Telefonaktiebolaget LM Ericsson
• Inventors: Dobler, Stefan
• Primary Examiner(s): Lerner; Martin
• Application Number:
  US10/496,769
• Publication Number:
  US 20050038652A1
• Time in Patent Office: 2,321 Days
• Field of Search: 704/243, 704/244, 704/248, 704/253, 704/255, 704/210, 704/215, 704/241
• US Class Current: 704/241
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 25/87 Detection of discrete points within a voice signal

Abstract

Method and device for the recognition of words and pauses in a voice signal. The words (W_i) spoken in a row and pauses (T_i) are thereby combined as to be appertaining to a word group as soon as one of the pauses (T_i) exceeds a limit value (TG). Stored references (R_j) are allocated to the voice signal of the word group, and an indication of the result of the allocation is effected after the limit value (TG) has been exceeded. To this end, parameters corresponding to the moments of the transitions between ranges with voice and non-voice are determined from the voice signal, and the limit value (TG) is then changed in dependence on said parameters.

Claims

1. A method for the recognition of words and pauses from a voice signal, wherein the consecutively spoken words (Wi) and pauses (Ti) are combined as to be appertaining to a word group as soon as one of the pauses (Ti) exceeds a limit value (TG), stored references (Rj) are allocated to the voice signal of the word group, and an indication of the result of the allocation is effected after the limit value (TG) has been exceeded, comprising the steps of:

determining the parameters corresponding to the moments of the transitions between ranges with voice and non-voice from the voice signal, and

changing the limit value (TG) in dependence on said parameters, wherein the parameters of the references (Rj) are the word beginning (RAj) and the word end (REj) of the stored references (R).

2. The method according to claim 1, wherein the moments of the transitions are the pause beginning (TAj) and the pause end (TEj) of the pauses (Ti) of the word group.

3. The method according to claim 2, further comprising the steps of:

increasing the limit value (TG) if at least for one pause (Ti) from the word group the time interval between the pause beginning (TAj) and the pause end (TEi) is longer than a pause time interval (TP), and

reducing the limit value (TG) if at least for one pause (Ti) from the word group the time interval between the pause beginning (TAj) and the pause end (TEi) is shorter than a pause time interval (TP).

4. The method for the recognition of words and pauses according to claim 3, further comprising the step of executing the method for each of the word groups in a number of consecutive word groups.

5. A method for the recognition of words and pauses from a voice signal, wherein the consecutively spoken words (Wi) and pauses (Ti) are combined as to be appertaining to a word group as soon as one of the pauses (Ti) exceeds a limit value (TG), stored references (Rj) are allocated to the voice signal of the word group, and an indication of the result of the allocation is effected after the limit value (TG) has been exceeded, comprising the steps of:

determining the parameters corresponding to the moments of the transitions between ranges with voice and non-voice from the voice signal, and

changing the limit value (TG) in dependence on said parameters, wherein the parameters of the references (Rj) are the word beginning (RAj) and the word end (REj) of the stored references (R), and further wherein the moments of the transitions are the word beginning (WA) and the word end (WEi) of the words (Wi) of the word group.

6. The method according to claim 5 further comprising the steps of:

increasing the limit value (TG), if at least for one word (Wi) from the word group the time interval between the word beginning (WAi) and the word end (WEi) is longer than the time interval between the word beginning (RAj) and the word end (REj) of the allocated reference (Rj), and

reducing the limit value (TG), if at least for one word (Wi) from the word group the time interval between the word beginning (WAi) and the word end (WEi) is shorter than the time interval between the word beginning (RAj) and the word end (REj) of the allocated reference (Rj).

7. The method according to claim 6, wherein the moments of the transitions are the pause beginning (TAj) and the pause end (TEj) of the pauses (Ti) of the word group.

8. The method according to claim 7, further comprising the steps of:

increasing the limit value (TG) if at least for one pause (Ti) from the word group the time interval between the pause beginning (TAj) and the pause end (TEi) is longer than a pause time interval (TP), and

reducing the limit value (TG) if at least for one pause (Ti) from the word group the time interval between the pause beginning (TAj) and the pause end (TEi) is shorter than a pause time interval (TP).

9. The method for the recognition of words and pauses according to claim 7, further comprising the step of executing the method for each of the word groups in a number of consecutive word groups.

10. The method according to claim 5, wherein the moments of the transitions are the pause beginning (TAj) and the pause end (TEj) of the pauses (Ti) of the word group.

11. The method according to claim 10, further comprising the steps of:

increasing the limit value (TG) if at east for one pause (Ti) from the word group the time interval between the pause beginning (TAj) and the pause end (TEi) is longer than a pause time interval (TP), and

reducing the limit value (TG) if at least for one pause (Ti) from the word group the time interval between the pause beginning (TAj) and the pause end (TEi) is shorter than a pause time interval (TP).

12. The method for the recognition of words and pauses according to claim 11, further comprising the step of executing the method for each of the word groups in a number of consecutive word groups.