SPEECH RECOGNITION APPARATUS AND SPEECH RECOGNITION METHOD

• US 20160372134A1
• Filed: 06/13/2016
• Published: 12/22/2016
• Est. Priority Date: 06/18/2015
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Nakadai, Kazuhiro, Mizumoto, Takeshi, Nakamura, Keisuke, Takigahira, Masayuki
• Granted Patent: US 9,697,832 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 2021/02166 Microphone arrays; Beamforming
• CPC Class Codes: G10L 21/0272 Voice signal separating
• CPC Class Codes: G10L 21/028 using properties of sound source
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10

Abstract

A speech recognition apparatus includes: a sound collection unit that collects a sound signal; a sound source localization unit that calculates a spatial spectrum from the sound signal that is collected by the sound collection unit and uses the calculated spatial spectrum to perform sound source localization; a speech zone determination unit that determines a zone in which a power of the spatial spectrum that is calculated by the sound source localization unit exceeds a predetermined threshold value based on a vehicle state; and a speech recognition unit that performs speech recognition with respect to a sound signal of the zone determined by the speech zone determination unit.

Claims

1. A speech recognition apparatus comprising:

a sound collection unit that collects a sound signal;

a sound source localization unit that calculates a spatial spectrum from the sound signal that is collected by the sound collection unit and uses the calculated spatial spectrum to perform sound source localization;

a speech zone determination unit that determines a zone in which a power of the spatial spectrum that is calculated by the sound source localization unit exceeds a predetermined threshold value based on a vehicle state; and

a speech recognition unit that performs speech recognition with respect to a sound signal of the zone determined by the speech zone determination unit.

2. The speech recognition apparatus according to claim 1, wherein

the vehicle state is at least one of a travel state of the vehicle and an operation state of the vehicle.

3. The speech recognition apparatus according to claim 1, wherein

a travel state of the vehicle is a state based on a vehicle speed, and

an operation state of the vehicle is at least one of an open and close state of a power window, an open and close state of a door, and an operation state of a wiper.

4. The speech recognition apparatus according to claim 1, comprising

a sound source separation unit that separates a speech signal of a driver seated on a driver seat from a speech signal of a passenger seated on a passenger seat based on the sound source localization result by the sound source localization unit, wherein

the speech recognition unit determines a zone in which a power of the spatial spectrum exceeds a predetermined threshold value based on the vehicle state with respect to the speech signal of the driver separated by the sound source separation unit or the speech signal of the passenger separated by the sound source separation unit.

5. The speech recognition apparatus according to claim 4, wherein

the predetermined threshold value is associated with a vehicle speed, an operation state of the vehicle, and a sound source of the driver or a sound source of the passenger.

6. The speech recognition apparatus according to claim 4, wherein

the predetermined threshold value is configured such that the predetermined threshold value with respect to a sound source of the driver is greater than the predetermined threshold value with respect to a sound source of the passenger.

7. A speech recognition method comprising:

(a) by way of a sound collection unit, collecting a sound signal;

(b) by way of a sound source localization unit, calculating a spatial spectrum from the sound signal that is collected in (a) and using the calculated spatial spectrum to perform sound source localization;

(c) by way of a speech zone determination unit, determining a zone in which a power of the spatial spectrum that is calculated in (b) exceeds a predetermined threshold value based on a vehicle state; and

(d) by way of a speech recognition unit, performing speech recognition with respect to a sound signal of the zone determined in (c).