WO2015196720A1 - 语音识别方法及*** - Google Patents
语音识别方法及*** Download PDFInfo
- Publication number
- WO2015196720A1 WO2015196720A1 PCT/CN2014/091947 CN2014091947W WO2015196720A1 WO 2015196720 A1 WO2015196720 A1 WO 2015196720A1 CN 2014091947 W CN2014091947 W CN 2014091947W WO 2015196720 A1 WO2015196720 A1 WO 2015196720A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- index
- recognition
- environment
- information
- environmental
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000004458 analytical method Methods 0.000 claims abstract description 66
- 230000007613 environmental effect Effects 0.000 claims description 146
- 230000005236 sound signal Effects 0.000 claims description 59
- 238000013507 mapping Methods 0.000 claims description 39
- 238000012545 processing Methods 0.000 claims description 18
- 230000000737 periodic effect Effects 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 7
- 230000004397 blinking Effects 0.000 description 6
- 239000003086 colorant Substances 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 230000001788 irregular Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 4
- 125000004122 cyclic group Chemical group 0.000 description 4
- 230000008859 change Effects 0.000 description 2
- 230000003749 cleanliness Effects 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
- 230000001755 vocal effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/20—Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech
Definitions
- speech recognition technology With the development of speech recognition technology, more and more devices (such as televisions, air conditioners, and other household appliances) can perform corresponding functions through voice control; for example, when a controlled device detects a voice control command, it can be based on The detected voice control instructions perform the corresponding operations.
- Speech recognition technology is easily affected by noisy vocals and other environmental noise, and the accuracy of speech recognition is directly related to environmental noise. For example, when the ambient noise is small, the speech recognition rate is high, but when the ambient noise is high or When a group of people around speak loudly, the speech recognition effect will become very poor, and even the controlled device cannot recognize the corresponding voice command.
- the human ear is not sensitive to environmental noise, and the human ear cannot accurately perceive the influence of environmental noise on speech recognition. Therefore, if the user issues a corresponding control command when it is not recognized, it will cause The problem that the control device cannot identify or cannot be accurately identified affects the operation of the controlled device.
- the step of acquiring the identification index corresponding to the environment recognition rate according to the analysis result of the sound information comprises:
- the step of displaying the prompt information of the identification index mapping according to the acquired identification index comprises:
- the corresponding indicator light is illuminated.
- the step of displaying the prompt information of the identification index mapping according to the acquired identification index comprises:
- the indicator column corresponding to the height is illuminated.
- the step of displaying the prompt information of the identification index mapping according to the acquired identification index comprises:
- the corresponding indicator light is gradually displayed according to the color corresponding to the three primary color values according to the found three primary color values.
- the step of analyzing the picked up sound information and acquiring the recognition index corresponding to the environment recognition rate according to the analysis result of the sound information comprises:
- the step of analyzing the picked up sound information and acquiring the recognition index corresponding to the environment recognition rate according to the analysis result of the sound information comprises:
- the environmental index is taken as the current identification index.
- the step of using the environmental index as the current identification index comprises:
- the first environmental index and the second environmental index are weighted and averaged to obtain a current environmental index; or, according to a mapping relationship between the preset feature information and the environmental index offset, determining the environmental index offset corresponding to the acquired feature information,
- the environmental bias adjusts the first environmental index to obtain a current identification index.
- the step of acquiring the characteristic information of the intermediate frequency signal in the signals of the respective preset frequencies is replaced by:
- the characteristic information of the intermediate frequency signal is acquired.
- the present invention also provides a voice recognition system, including:
- a voice pickup module for picking up sound information from the environment
- a noise analysis module configured to analyze the picked up sound information, and obtain a recognition index corresponding to the environment recognition rate according to the analysis result of the sound information
- the information display module is configured to display prompt information of the identification index map according to the obtained identification index, and the user performs a corresponding operation according to the prompt information.
- the noise analysis module is further configured to:
- the information display module is further configured to:
- the corresponding indicator light is illuminated.
- the information display module is further configured to:
- the indicator column corresponding to the height is illuminated.
- the information display module is further configured to:
- the corresponding indicator light is gradually displayed according to the color corresponding to the three primary color values according to the found three primary color values.
- the noise analysis module comprises:
- a voice index acquiring unit configured to analyze the picked up sound information, and obtain a voice recognition index corresponding to the environment recognition rate according to the analysis result of the sound information
- a distance index obtaining unit configured to measure a distance between a controlled device that picks up the sound information and a sound source that sends a voice command to control the controlled device, obtain a corresponding distance parameter, and obtain the Distance identification index corresponding to the distance parameter;
- a processing unit configured to calculate, according to the acquired speech recognition index and the distance recognition index, a comprehensive recognition index corresponding to the acquired environment recognition rate according to a preset rule, and use the comprehensive recognition index as the current identification index.
- the noise analysis module comprises:
- An analyzing unit configured to analyze an intensity or a ratio of an audio signal of each preset frequency in the picked up sound information
- a determining unit configured to determine, according to a preset mapping relationship between an intensity or a ratio of the audio signal of each frequency and an environment index, a first environmental index corresponding to the obtained intensity or proportion of the audio signal of each preset frequency;
- a processing unit configured to use the environmental index as a current identification index.
- the processing unit comprises:
- Obtaining a subunit configured to acquire feature information of an intermediate frequency signal among signals of respective preset frequencies
- Determining a subunit configured to determine, according to a mapping relationship between the preset feature information and the second environment index, a second environment index corresponding to the acquired feature information, and according to the preset first environment index and the second environment index Weighting, the weighted average of the first environment index and the second environment index to obtain a current environment index; or, according to a mapping relationship between preset feature information and an environmental index offset, determining the acquired feature information corresponding to The environmental index is biased, and the first environmental index is adjusted by the environmental bias to obtain a current identification index.
- the acquiring subunit is further configured to: when the acquired first environmental index is greater than the first preset threshold, or the signals of the respective preset frequencies, the frequency or intensity of the intermediate frequency signal is greater than the second preset threshold And acquiring feature information of the intermediate frequency signal.
- the controlled device picks up the sound information from the environment; analyzes the picked up sound information, and obtains a recognition index corresponding to the environment recognition rate according to the analysis result of the sound information; and acquires according to the noise analysis module.
- the identification index displays the prompt information of the identification index mapping, and the user performs a corresponding operation according to the prompt information; so that the controlled device can display the environmental noise situation recognized by the user in time, prompting the user to the above environment.
- the noise situation is convenient for the user to take corresponding measures according to the environmental noise situation prompted by the controlled device; the interaction of the human-machine is improved.
- FIG. 1 is a schematic diagram of a hardware architecture of a preferred embodiment of a controlled device for implementing voice recognition according to the present invention
- FIG. 2 is a schematic diagram of functional modules of a preferred embodiment of the speech recognition system of the present invention.
- FIG. 1 is a schematic diagram of a hardware architecture of a preferred embodiment of a controlled device for implementing voice recognition according to the present invention.
- the controlled device 100 includes a processing unit 11, a storage unit 12, a voice receiving unit 13, and a voice recognition system 14.
- the controlled device 100 includes, but is not limited to, a home appliance (such as an air conditioner or the like), a mobile terminal (such as a mobile phone, a tablet computer, etc.), a remote controller, or any other suitable electronic device (preferably an air conditioner).
- the storage unit 12 is configured to store the voice recognition system 14 and its operation data, and the sound information picked up by the voice recognition system 14, the recognition index of the environment recognition rate mapped with the picked up sound information, and the like. It should be emphasized that the storage unit 12 may be a single storage device or a collective name of a plurality of different storage devices, and details are not described herein.
- the voice receiving unit 13 is configured to: under the control of the processing unit 11, detect whether the user sends a voice instruction, and after detecting that the user sends the voice command, pick up the voice information in the environment; or, in the process Under the control of the unit 11, the sound information in the environment is picked up in real time or according to a preset period.
- the processing unit 11 is configured to invoke and execute the voice recognition system 14 to pick up sound information from the environment; after picking up the corresponding sound information, analyzing the picked up sound information, and according to the Obtaining an identification index corresponding to the environment recognition rate, and displaying the identification information of the identification index map according to the acquired identification index, so that the user performs a corresponding operation according to the prompt information, thereby implementing the controlled
- the device 100 can prompt the user according to the identified environmental noise condition, and the user can adopt a corresponding processing strategy according to the environmental noise situation prompted by the controlled device 100.
- the processing unit 11 and the storage unit 12 may be separate units, or may be integrated to form a controller, which is not described herein.
- the sound information can be understood as the sound information corresponding to the voice control command sent by the user;
- the prompt information can be understood as: the mapped index according to the difference of the recognition index of the environment recognition rate corresponding to the sound information. Different information; for example, the higher the recognition index corresponding to the environment recognition rate, indicating that the lower the noise in the environment is, the corresponding prompt information is also different; for example, when the recognition index is advanced, the green indication is displayed. Light; when the recognition index is intermediate, the yellow indicator light is displayed; when the recognition index is low, the red indicator light is displayed, prompting the user to send the corresponding strategy to be used when the corresponding voice command is sent.
- the prompt information may also be different according to specific hardware devices, production costs, usage environments, etc., and the specific expressions thereof may also be different, and details are not described herein.
- FIG. 2 is a schematic diagram of functional modules of a preferred embodiment of the speech recognition system of the present invention.
- the functional block diagram shown in FIG. 2 is merely an exemplary diagram of a preferred embodiment, and the function of the voice recognition system 14 shown in FIG. 2 by those skilled in the art.
- Modules which can be easily supplemented by new functional modules; the names of the functional modules are custom names, which are only used to assist in understanding the various program function blocks of the speech recognition system 14, and are not used to define the technical solution of the present invention.
- the function to be achieved by the functional modules that each define the name At the heart of the technical solution is the function to be achieved by the functional modules that each define the name.
- the voice recognition system 14 of the present embodiment runs in the controlled device 100 of FIG. 1.
- the voice recognition system 14 includes: a voice pickup module 141, a noise analysis module 142, and an information display module 143. Functions of the modules as follows:
- a voice picking module 141 configured to pick up sound information from the environment
- the sound information picked up by the voice picking module 141 from the current environment includes noise of the current environment.
- the voice pickup module 141 picks up the corresponding sound information from the current environment in real time. In order to save the running cost of the voice recognition system 14 and reduce the power consumption of the voice recognition system 14, the voice picking module 141 can also pick up the corresponding voice information from the current environment according to a preset period.
- the preset period may be customized by the voice recognition system 14, for example, setting the preset period to 10 ms, or may be set by the user according to his own needs, or when the controlled device 100 running the voice recognition system 14 is shipped from the factory, The setting is performed by the technician according to the operating environment of the voice recognition system 14; the specific duration of the preset period is not specifically limited in this embodiment. Of course, the preset period can also be updated or reset according to specific needs.
- the noise analysis module 142 is configured to analyze the picked up sound information, and obtain an identification index corresponding to the environment recognition rate according to the analysis result of the sound information;
- the noise analysis module 142 analyzes the sound information picked up by the voice picking module 141, for example, acquires the sound decibel corresponding to the noise contained in the sound information, or It analyzes whether the picked up sound information contains a voice command sent by the user. Based on the analysis result of the sound information, the noise analysis module 142 acquires the recognition index corresponding to the current environment recognition rate.
- the recognition index corresponding to the environment recognition rate can be understood as: the voice recognition system 14 can recognize the recognition ability of the valid voice command sent by the user; the higher the recognition index, the current environment of the voice recognition system 14 The smaller the noise in the speech, the higher the recognition rate of the voice command sent by the speech recognition system 14 to the user.
- the information display module 143 is configured to display prompt information of the identification index map according to the acquired identification index, and the user performs a corresponding operation according to the prompt information.
- the information display module 143 displays the identification index according to the recognition index acquired by the noise analysis module 142.
- the prompt information of the index display is displayed, and the user performs a corresponding operation according to the prompt information displayed by the information display module 143.
- the user selects whether to issue a corresponding voice command according to the prompt, or whether the user needs to send a corresponding voice command in a loud voice. Or whether the user needs to reduce the noise in the current environment to take a normal voice decibel to issue a corresponding voice command or the like; for example, if the noise of the current environment corresponding to the prompt information displayed by the information display module 143 is low, In the current environment, the influence of the environmental noise on the voice recognition system 14 to recognize the voice command is almost negligible, that is, the user can send the corresponding voice command in an ordinary manner, and the user according to the prompt information of the information display module 143, according to the ordinary The way is to issue the corresponding voice command.
- the voice picking module of the voice recognition system of the embodiment of the present invention picks up sound information from the environment; the noise analysis module analyzes the sound information picked up by the voice picking module, and according to the analysis result of the sound information, the noise analysis module acquires the environment identifier.
- the information display module displays the prompt information of the identification index map according to the identification index acquired by the noise analysis module, and the user performs a corresponding operation according to the prompt information; so that the voice recognition system can
- the identified environmental noise situation is displayed in time, prompting the user to the above-mentioned environmental noise situation, and is convenient for the user to take corresponding measures according to the environmental noise situation prompted by the voice recognition system; improving the intelligence of the voice recognition system and the interaction of the human-machine Further, since the user can specifically transmit the corresponding voice command according to the prompted environmental noise condition, the recognition accuracy of the voice recognition system for the voice command is improved.
- the first embodiment when the noise analysis module 142 analyzes the noise decibel corresponding to the sound information picked up by the voice pickup module 141, the noise analysis module 142 searches for a noise recognition strategy table, from the noise recognition strategy table, the noise analysis module 142. Acquire the identification index of the noise decibel map. Further, the information display module 143 displays the corresponding prompt information according to the recognition index acquired by the noise analysis module 142.
- the noise analysis module 142 includes:
- a voice index acquiring unit configured to analyze the picked up sound information, and obtain a voice recognition index corresponding to the environment recognition rate according to the analysis result of the sound information
- the voice index acquisition unit analyzes the picked up sound information, for example, acquires a sound decibel corresponding to the noise contained in the sound information, or analyzes the picked up sound information. Whether it contains a voice command sent by the user.
- the speech index acquisition unit acquires a speech recognition index corresponding to the current environment recognition rate according to the analysis result of the sound information.
- the voice recognition index corresponding to the environment recognition rate can be understood as: the voice recognition system 14 can recognize the recognition capability of the valid voice command sent by the user; the higher the voice recognition index, the current voice recognition system 14 The smaller the noise in the environment, the higher the recognition rate of the voice command sent by the speech recognition system 14 to the user.
- a distance index obtaining unit configured to measure a distance between a controlled device that picks up the sound information and a sound source that sends a voice command to control the controlled device, obtain a corresponding distance parameter, and obtain the Distance identification index corresponding to the distance parameter;
- the distance index acquisition unit measures: the distance between the controlled device that picks up the sound information, that is, the controlled device 100 described in FIG. 1 and the sound source that sends the voice command, to obtain a corresponding distance parameter.
- the distance index obtaining unit measures a distance recognition index mapped with the distance parameter according to the distance parameter; the distance recognition index can be understood as: indicating a sound source for transmitting a valid voice command.
- the distance between the controlled devices 100 is far, that is, the closer the sound source is to the controlled device 100, the higher the distance recognition index; the sound source and the controlled device 100 The further the distance, the lower the distance recognition index.
- the higher the distance recognition index the higher the recognition rate of the voice command by the controlled device 100.
- the distance index obtaining unit may measure the distance by using an infrared technology, a distance sensor, or an NFC technology, etc., when the distance index obtaining unit measures the distance between the user and the controlled device 100, and the embodiment of the present invention
- the specific manner in which the distance obtaining unit obtains the distance parameter is not limited.
- the distance between the user and the controlled device 100 measured by the distance index obtaining unit may be understood as: a horizontal distance between the user center point and the controlled device 100;
- the distance parameter acquired by the distance index obtaining unit may be a horizontal distance between a sound source that sends a voice instruction and the controlled device 100; the sound source may be understood as a user, a speaker or a speaker that sends a valid voice command. other devices.
- the preferred sound source is a user.
- a processing unit configured to calculate, according to the acquired speech recognition index and the distance recognition index, a comprehensive recognition index corresponding to the acquired environment recognition rate according to a preset rule, and use the comprehensive recognition index as the current identification index.
- the processing unit calculates the comprehensive recognition index corresponding to the acquired environment recognition rate according to the preset rule according to the speech recognition index acquired by the speech index acquisition unit and the distance recognition index acquired by the distance index acquisition unit.
- the preset rule may be understood as: a calculation rule adopted by the controlled device 100 by default, or a calculation rule set by the technician for the controlled device 100 according to a specific use scenario; by default, the comprehensive recognition index is a voice.
- the sum of the identification index and the distance recognition index For example, in the current environment, the ambient noise corresponds to a speech recognition index of 52, and the corresponding distance recognition index is -5, and the sum of the speech recognition index and the distance recognition index is 47, that is, the comprehensive recognition index is 47.
- the noise analysis module 142 includes:
- An analyzing unit configured to analyze an intensity or a ratio of an audio signal of each preset frequency in the picked up sound information
- the strength or proportion of the audio signal of each preset frequency in the extracted sound information obtained by the analyzing unit can be implemented by:
- the first audio signal, the second audio signal, and the third audio signal of HZ ⁇ 4KHZ and 4KHZ ⁇ 20KHZ, the amplitudes a, b, and c of the first audio signal, the second audio signal, and the third audio signal are first
- the intensity of the audio signal, the second audio signal, and the third audio signal, and the frequencies of the first audio signal, the second audio signal, and the third audio signal are respectively a/(a+b+c), b/(a+b +c), c/(a+b+c).
- a determining unit configured to determine, according to a preset mapping relationship between an intensity or a ratio of the audio signal of each frequency and an environment index, a first environmental index corresponding to the obtained intensity or proportion of the audio signal of each preset frequency;
- the mapping relationship between the strength or the ratio of the audio signal of each preset frequency and the environment index may be a mapping relationship between the strength or the ratio of the audio signal of each frequency and the environment index, or according to The formula of the environmental index of the intensity or ratio of the audio signal of each frequency, for example, the preset respective frequencies are low frequency, intermediate frequency and high frequency, for example, the low frequency signal, the intermediate frequency signal, and the high frequency signal have lower intensities, then corresponding
- the environmental index is 30, and when the strength of the low frequency signal and the high frequency signal is low, and the intensity of the high frequency signal is high, the environmental index is 45 at this time, and the environmental index corresponds to the severity of the environment; and in other embodiments
- the environmental index can also be the opposite of the current environmental cleanliness.
- a processing unit configured to use the environmental index as a current identification index.
- the processing unit includes: an acquiring subunit, configured to acquire feature information of the intermediate frequency signal in the signals of the respective preset frequencies;
- Determining a subunit configured to determine, according to a mapping relationship between the preset feature information and the second environment index, a second environment index corresponding to the acquired feature information, and according to the preset first environment index and the second environment index Weighting, the weighted average of the first environment index and the second environment index to obtain a current environment index; or, according to a mapping relationship between preset feature information and an environmental index offset, determining the acquired feature information corresponding to The environmental index is biased, and the first environmental index is adjusted by the environmental bias to obtain a current identification index.
- the feature information may include a pulse signal, a periodic cycle signal, and the intensity is substantially constant, and the irregular signal is strong and the intensity is high.
- the characteristic information of the intermediate frequency signal is a pulse, it indicates that the intermediate frequency signal is a human voice interference, which is extremely unfavorable for identification, and the second environmental index is higher as the second environmental index is 80 (the second environmental index represents the environmental severity);
- the characteristic information of the intermediate frequency signal is a periodic cyclic signal and the intensity is substantially constant
- the intermediate frequency signal is a fixed interference generated by a fixed device such as an exhaust fan, and such interference is easily filtered, corresponding to a second environmental index such as a second environmental index of 65.
- the second environmental index represents the severity of the environment
- the characteristic information of the intermediate frequency signal is an irregular signal and the intensity is large, indicating that the intermediate frequency signal is a sound source that is many and complicated, and is not easy to be filtered out, corresponding to the second environmental index being high
- the manufacturer may preset a mapping relationship between the feature information and the second environment index, and invoke the corresponding second environment according to the acquired feature information.
- the second environmental index corresponding to the intensity of the pulse signal in the characteristic information may also be different, the periodic cycle signal A second environment index corresponding to the different intensities or different.
- the corresponding environmental index offset can be obtained based on the characteristic information of the intermediate frequency signal, and the first environmental index is adjusted based on the environmental index offset to obtain Current environmental index.
- the characteristic information of the intermediate frequency signal is a pulse, it indicates that the intermediate frequency signal is a human voice interference, which is extremely unfavorable for identification, and the corresponding environmental index offset is medium, such as an environmental index offset 3; the characteristic information of the intermediate frequency signal is a periodic cyclic signal and the intensity
- the IF signal is a fixed interference emitted by a fixed device such as an exhaust fan.
- the corresponding environmental index offset is small, such as an environmental index offset of -1; the characteristic information of the intermediate frequency signal is an irregular signal.
- the intensity is large, the IF signal is more and more complex, and it is not easy to filter out.
- the corresponding environmental index is large.
- the environmental index is less than 5, and the manufacturer can preset the characteristic information and the environmental index offset.
- the mapping relationship, and according to the acquired feature information, the corresponding environmental index offset is called, and the environmental index offset corresponding to the intensity of the pulse signal in the feature information may also be different, and the environmental index offset corresponding to the intensity of the periodic cycle signal is different. It can also be different.
- the acquiring subunit is further configured to: when the obtained first environment index is greater than a first preset threshold, or a signal of each preset frequency, a frequency of the intermediate frequency signal Or the characteristic information of the intermediate frequency signal is obtained when the intensity is greater than the second preset threshold.
- the third acquiring module is The feature information of the IF signal is obtained, so that the second determining module does not need to determine the second environment index or the environmental index offset corresponding to the feature information each time the audio signal is detected, thereby saving system energy consumption.
- the processing unit is further configured to: when the obtained environmental index is less than or equal to the first preset threshold, or the signal of each preset frequency, the frequency or intensity of the intermediate frequency signal is less than or equal to the second preset threshold
- the first environmental index is taken as the current identification index.
- the three methods for obtaining the recognition index corresponding to the environment recognition rate listed above are merely exemplary. Those skilled in the art can use the technical idea of the present invention to obtain other recognition indexes corresponding to the environmental recognition rate according to their specific needs. The manner of the present invention is all within the scope of the present invention and is not exhaustive.
- the information display module 143 may display the prompt information of the identification index mapping according to the identification index acquired by the noise analysis module 142 in the following manner:
- the noise of the different decibels is displayed by using the color of the indicator light. For example, if the noise decibel obtained by the noise analysis module 142 is less than 50 db (decibel), the information display module 143 displays a green light; the noise analysis module 142 obtains The information display module 143 displays a yellow light when the noise decibel is between 50db and 70db (including 50db, including 70db); the noise decibel obtained by the noise analysis module 142 is 70db or more (including 70db). The information display module 143 displays a red light. For example, if the noise analysis module 142 analyzes that the noise decibel corresponding to the sound information picked up by the voice pickup module 141 is 45 db, the information display module 143 displays a green light.
- the height of the indicator column is used to display the noise of different decibels, wherein the indicator column can be arranged horizontally or vertically by a plurality of display units, and the number of display units to be displayed is calculated according to different noise decibels, thereby correspondingly displaying The height of the indicator post.
- the indicator columns composed of 10 display units are vertically arranged.
- the information display module 143 lights up a display unit at the bottom of the indicator column.
- the information display module 143 illuminates all the 10 display units from the lowest to the top of the indicator column, and the number of display units Corresponding to the height of the indicator column to remind the user of the noise level of the environment. For example, if the noise decibel of the current environment acquired by the noise analysis module 142 is 65 db, the information display module 143 illuminates 8 display units.
- the lamp post seen by the user is relatively high, it is necessary to use a relatively large lamp column. The sound is such that the speech recognition system 14 is able to accurately identify the corresponding speech command.
- the three primary colors of the indicator light are used to display the noise of different decibels, and the change of the magnitude of the noise decibel value in the environment is converted into the corresponding components of the three primary colors of RGB, and the brightness values of the primary color components in the three primary colors are adjusted according to the noise decibel value, thereby matching A gradient is produced, and the gradient of the display color is performed as the value of the noise decibel value.
- the noise decibel value obtained by the noise analysis module 142 is also increased, and the color value of the information display module 143 according to the noise decibel value is also determined by the light color system. Gradually changing to the dark color, prompting users to become more and more noisy.
- the noise of the indicator light is used to display the noise of different decibels.
- the indicator light displayed by the information display module 143 is The light is always on, that is, the indicator light that is seen by the human eye is not blinking; as the noise is getting larger, the blinking frequency of the indicator light displayed by the information display module 143 is also higher and higher, thereby prompting the user. The noise is getting bigger and bigger. It can also be understood that the blinking frequency of the indicator light displayed by the information display module 143 is proportional to the magnitude of the ambient noise.
- the speech recognition system of the present invention improves the intelligence of the speech recognition system by displaying the recognized noise to prompt the user to the noise situation in the current environment; further, because the environmental noise condition is prompted to be correspondingly transmitted correspondingly The voice command, therefore, improves the recognition accuracy of the voice recognition system for voice commands.
- FIG. 3 is a schematic flowchart of a voice recognition method according to a preferred embodiment of the present invention.
- the embodiment of the present invention provides an embodiment of a voice recognition method.
- the speech recognition method of the present invention includes the following steps:
- Step S01 picking up sound information from the environment
- the sound information picked up by the controlled device from the current environment includes noise of the current environment.
- the controlled device picks up the corresponding sound information from the current environment in real time.
- the controlled device can also pick up the corresponding sound information from the current environment according to a preset period.
- the preset period may be customized by the controlled device, for example, the preset period is set to 10 ms, or may be set by the user according to his own needs, or the technician may operate according to the controlled device when the controlled device is shipped from the factory.
- the environment is set; the specific duration of the preset period is not specifically limited in this embodiment. Of course, the preset period can also be updated or reset according to specific needs.
- Step S02 analyzing the picked up sound information, and acquiring an identification index corresponding to the environment recognition rate according to the analysis result of the sound information;
- the controlled device When the corresponding sound information is picked up from the environment, the controlled device analyzes the picked up sound information, for example, acquires a sound decibel corresponding to the noise contained in the sound information, or analyzes whether the picked up sound information is Contains voice commands sent by the user. According to the analysis result of the sound information, the controlled device acquires a recognition index corresponding to the current environment recognition rate.
- the identification index corresponding to the environment recognition rate can be understood as: the controlled device can recognize the recognition capability of the valid voice command sent by the user; the higher the recognition index, indicates that the controlled device is currently in the environment. The smaller the noise, the higher the recognition rate of the voice command sent by the controlled device to the user.
- Step S03 Display, according to the acquired identification index, prompt information of the identification index mapping, for the user to perform a corresponding operation according to the prompt information.
- the controlled device After the controlled device obtains the identification index corresponding to the environment recognition rate according to the analysis result of the picked up sound information, the controlled device displays the prompt information of the identification index map according to the acquired identification index, for the user to The prompt information displayed by the controlled device performs a corresponding operation.
- the user selects whether to issue a corresponding voice instruction according to the prompt, or whether the user needs to issue a corresponding voice instruction in a loud voice, or Whether the user needs to reduce the noise in the current environment to take a normal voice decibel to issue a corresponding voice command or the like; for example, according to the noise of the current environment corresponding to the prompt information displayed by the controlled device, the current environment is The impact of the ambient noise on the voice command of the controlled device is almost negligible, that is, the user can send the corresponding voice command in an ordinary manner, and the user sends the corresponding message according to the prompt information of the controlled device according to the normal manner. Voice command.
- the controlled device picks up the sound information from the environment; analyzes the picked up sound information, and obtains a recognition index corresponding to the environment recognition rate according to the analysis result of the sound information;
- the identification index is displayed, and the prompt information of the identification index mapping is displayed, and the user performs a corresponding operation according to the prompt information; so that the controlled device can display the environmental noise condition recognized by the user in time, and prompt the user to the above-mentioned environmental noise.
- the situation is convenient for the user to take corresponding measures according to the environmental noise situation prompted by the controlled device; the interactivity of the human-machine is improved; further, the user can specifically send the corresponding voice command according to the prompted environmental noise condition. Therefore, the recognition accuracy of the voice command by the controlled device is improved.
- the controlled device obtains the recognition index corresponding to the environment recognition rate according to the analysis result of the sound information, and may adopt the following manner:
- the first embodiment is: when the controlled device analyzes the noise decibel corresponding to the picked up sound information, the noise identification strategy table is searched, and the recognition index of the noise decibel map is obtained from the noise recognition strategy table. Further, according to the acquired identification index, the controlled device displays corresponding prompt information.
- step S02 includes:
- the controlled device When the corresponding sound information is picked up from the environment, the controlled device analyzes the picked up sound information, for example, acquires a sound decibel corresponding to the noise contained in the sound information, or analyzes whether the picked up sound information is Contains voice commands sent by the user. According to the analysis result of the sound information, the controlled device acquires a voice recognition index corresponding to the current environment recognition rate.
- the voice recognition index corresponding to the environment recognition rate can be understood as: the voice recognition system 14 can recognize the recognition capability of the valid voice command sent by the user; the higher the voice recognition index, the current voice recognition system 14 The smaller the noise in the environment, the higher the recognition rate of the voice command sent by the speech recognition system 14 to the user.
- the controlled device acquires the distance between the sound source and the sound source that sends the voice command to obtain a corresponding distance parameter. After obtaining the corresponding distance parameter, the controlled device measures a distance recognition index mapped to the distance parameter according to the distance parameter; the distance recognition index can be understood as: indicating a sound source and a place for transmitting an effective voice command.
- the distance between the controlled devices is far, that is, the closer the sound source is to the controlled device, the higher the distance recognition index; the farther the sound source is from the controlled device, Then the distance recognition index is lower. The higher the distance recognition index, the higher the recognition rate of the voice command by the controlled device.
- the measurement of the distance may be implemented by using an infrared technology, a distance sensor, or an NFC technology.
- the embodiment of the present invention acquires the The specific method of the distance parameter is not limited.
- the distance between the controlled device and the user may be understood as: a horizontal distance between the user center point and the controlled device; in general, the distance parameter may be set to send a voice.
- the horizontal distance between the commanded sound source and the controlled device; the sound source can be understood as a user, speaker or other device that transmits a valid voice command.
- the preferred sound source is a user.
- a processing unit configured to calculate, according to the acquired speech recognition index and the distance recognition index, a comprehensive recognition index corresponding to the acquired environment recognition rate according to a preset rule, and use the comprehensive recognition index as the current identification index.
- the controlled terminal calculates a comprehensive recognition index corresponding to the environment recognition rate according to the voice recognition index and the distance recognition index acquired by the distance index acquisition unit according to a preset rule.
- the preset rule may be understood as: a calculation rule adopted by the controlled device by default, or a calculation rule set by the technician for the controlled device according to the specific use scenario; by default, the comprehensive recognition index is a speech recognition index.
- the sum of the distance recognition index For example, in the current environment, the ambient noise corresponds to a speech recognition index of 52, and the corresponding distance recognition index is -5, and the sum of the speech recognition index and the distance recognition index is 47, that is, the comprehensive recognition index is 47.
- the environmental index is taken as the current identification index.
- acquiring the intensity or proportion of the audio signal of each preset frequency in the detected mixed audio signal can be implemented by:
- the detected mixed audio signals are separately filtered by bandpass filters of different frequencies to obtain audio signals of different frequencies, for example, the preset frequencies are 20HZ ⁇ 300HZ, 300 respectively.
- HZ ⁇ 4KHZ and 4KHZ ⁇ 20KHZ, 20HZ ⁇ 300HZ, 300 can be obtained by bandpass filter
- the first audio signal, the second audio signal, and the third audio signal of HZ ⁇ 4KHZ and 4KHZ ⁇ 20KHZ, the amplitudes a, b, and c of the first audio signal, the second audio signal, and the third audio signal are first
- the intensity of the audio signal, the second audio signal, and the third audio signal, and the frequencies of the first audio signal, the second audio signal, and the third audio signal are respectively a/(a+b+c), b/(a+b +c), c/(a+b+c).
- the mapping relationship between the intensity or the ratio of the audio signal of each preset frequency and the environment index may be a mapping relationship between the strength or the ratio of the audio signal of each frequency and the environment index, or Calculating the formula of the environmental index according to the intensity or ratio of the audio signal of each frequency, for example, the preset respective frequencies are low frequency, intermediate frequency, and high frequency, for example, the low frequency signal, the intermediate frequency signal, and the high frequency signal have lower intensities, corresponding to The environmental index is 30, and when the strength of the low frequency signal and the high frequency signal is low, and the intensity of the high frequency signal is high, the environmental index is 45 at this time, and the environmental index corresponds to the severity of the environment; and in other embodiments In the middle, the environmental index can also be the opposite of the current environmental cleanliness.
- the step of using the environmental index as the current identification index includes:
- the first environmental index and the second environmental index are weighted and averaged to obtain a current environmental index; or, according to a mapping relationship between the preset feature information and the environmental index offset, determining the environmental index offset corresponding to the acquired feature information,
- the environmental bias adjusts the first environmental index to obtain a current identification index.
- the feature information may include a pulse signal, a periodic cycle signal, and the intensity is substantially constant, and the irregular signal is strong and the intensity is high.
- the characteristic information of the intermediate frequency signal is a pulse, it indicates that the intermediate frequency signal is a human voice interference, which is extremely unfavorable for identification, and the second environmental index is higher as the second environmental index is 80 (the second environmental index represents the environmental severity);
- the characteristic information of the intermediate frequency signal is a periodic cyclic signal and the intensity is substantially constant
- the intermediate frequency signal is a fixed interference generated by a fixed device such as an exhaust fan, and such interference is easily filtered, corresponding to a second environmental index such as a second environmental index of 65.
- the second environmental index represents the severity of the environment
- the characteristic information of the intermediate frequency signal is an irregular signal and the intensity is large, indicating that the intermediate frequency signal is a sound source that is many and complicated, and is not easy to be filtered out, corresponding to the second environmental index being high
- the manufacturer may preset a mapping relationship between the feature information and the second environment index, and invoke the corresponding second environment according to the acquired feature information.
- the second environmental index corresponding to the intensity of the pulse signal in the characteristic information may also be different, the periodic cycle signal A second environment index corresponding to the different intensities or different.
- the corresponding environmental index offset can be obtained based on the characteristic information of the intermediate frequency signal, and the first environmental index is adjusted based on the environmental index offset to obtain Current environmental index.
- the characteristic information of the intermediate frequency signal is a pulse, it indicates that the intermediate frequency signal is a human voice interference, which is extremely unfavorable for identification, and the corresponding environmental index offset is medium, such as an environmental index offset 3; the characteristic information of the intermediate frequency signal is a periodic cyclic signal and the intensity
- the IF signal is a fixed interference emitted by a fixed device such as an exhaust fan.
- the corresponding environmental index offset is small, such as an environmental index offset of -1; the characteristic information of the intermediate frequency signal is an irregular signal.
- the intensity is large, the IF signal is more and more complex, and it is not easy to filter out.
- the corresponding environmental index is large.
- the environmental index is less than 5, and the manufacturer can preset the characteristic information and the environmental index offset.
- the mapping relationship, and according to the acquired feature information, the corresponding environmental index offset is called, and the environmental index offset corresponding to the intensity of the pulse signal in the feature information may also be different, and the environmental index offset corresponding to the intensity of the periodic cycle signal is different. It can also be different.
- the step of acquiring the characteristic information of the intermediate frequency signal in the signals of the respective preset frequencies is replaced by:
- the characteristic information of the intermediate frequency signal is acquired.
- the third acquiring module is The feature information of the IF signal is obtained, so that the second determining module does not need to determine the second environment index or the environmental index offset corresponding to the feature information each time the audio signal is detected, thereby saving system energy consumption.
- the processing unit is further configured to: when the obtained environmental index is less than or equal to the first preset threshold, or the signal of each preset frequency, the frequency or intensity of the intermediate frequency signal is less than or equal to the second preset threshold
- the first environmental index is taken as the current identification index.
- the three methods for obtaining the recognition index corresponding to the environment recognition rate listed above are merely exemplary. Those skilled in the art can use the technical idea of the present invention to obtain other recognition indexes corresponding to the environmental recognition rate according to their specific needs. The manner of the present invention is all within the scope of the present invention and is not exhaustive.
- the controlled device displays the prompt information of the identification index mapping according to the obtained identification index, and may adopt the following manner:
- the controlled device uses the color of the indicator light to display noise of different decibels. For example, if the noise decibel obtained by the controlled device is less than 50 db (decibel), the controlled device displays a green light; the controlled device acquires The noise decibel is between 50db ⁇ 70db (including 50db, excluding 70db), then the controlled device displays a yellow light; the noise decibel obtained by the controlled device is 70db or more (including 70db), then The controlled device displays a red light. For example, if the controlled device analyzes that the noise information corresponding to the sound information picked up by the voice pickup module 141 is 45 db, the controlled device displays a green light.
- the controlled device uses the height of the indicator bar to display noise of different decibels, wherein the indicator column can be arranged horizontally or vertically by a plurality of display units, and the number of display units to be displayed is calculated according to different noise decibels. Therefore, the height of the indicator column is displayed correspondingly.
- the indicator columns composed of 10 display units are vertically arranged.
- the noise decibel obtained by the controlled device is 40 db
- the controlled device illuminates a display unit at the bottom of the indicator column
- the noise decibel obtained by the controlled device is 70 db
- the controlled device illuminates all 10 display units from the lowest to the top of the indicator column, and the number of display units corresponds to the indicator column.
- the height is to alert the user to the noise level of the environment.
- the controlled device illuminates 8 display units, and when the lamp post seen by the user is relatively high, it is known that a relatively large sound is needed. In order to enable the controlled device to accurately identify the corresponding voice instruction.
- the three primary colors of the indicator light are used to display the noise of different decibels, and the change of the magnitude of the noise decibel value in the environment is converted into the corresponding components of the three primary colors of RGB, and the brightness values of the primary color components in the three primary colors are adjusted according to the noise decibel value, thereby matching A gradient is produced, and the gradient of the display color is performed as the value of the noise decibel value.
- the noise decibel value obtained by the controlled device also becomes larger, and the color value of the controlled device according to the noise decibel value is also gradually changed from the light color system. The dark color changes, prompting the user to get louder and louder.
- the indicator light displayed by the controlled device is always on That is to say, the indicator light seen by the human eye does not blink; as the noise becomes larger, the blinking frequency of the indicator light displayed by the controlled device is also higher and higher, thereby prompting the user that the noise is coming. The bigger. It can also be understood that the blinking frequency of the indicator light displayed by the controlled device is proportional to the magnitude of the ambient noise.
- the controlled device of the present invention improves the intelligence of the controlled device by displaying the identified noise to prompt the user to the noise situation in the current environment; further, the corresponding voice can be specifically transmitted according to the prompted environmental noise condition.
- the instruction therefore, improves the accuracy of recognition of voice commands by the controlled device.
- the foregoing embodiment method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be through hardware, but in many cases, the former is better.
- Implementation Based on such understanding, the technical solution of the present invention, which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM, disk,
- the optical disc includes a number of instructions for causing a terminal device (which may be a cell phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the methods described in various embodiments of the present invention.
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
一种语音识别方法和***,该方法包括:从环境中拾取出声音信息(S01);分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数(S02);根据噪音分析模块获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作(S03)。该语音识别方法和***使得受控设备能够将自身识别出的环境噪声情况及时显示出来,提示用户上述环境噪声情况,便于用户根据所述受控设备提示的环境噪声情况,采取相应措施,提高了人机的可交互性。
Description
技术领域
本发明涉及语音识别技术,尤其涉及一种语音识别方法及***。
背景技术
随着语音识别技术的发展进步,越来越多的设备(比如电视机、空调器等家用电器)可通过语音控制来执行相应的功能;比如,受控设备检测到语音控制指令时,可以根据检测到的语音控制指令来执行相应的操作。而语音识别技术很容易受到嘈杂的人声以及其他环境噪音的影响,且语音识别的准确率直接与环境噪音相关,比如当环境噪音较小时,语音识别率较高,但是当环境噪音很大或者周围一群人高声说话时,语音识别效果将会变得很差,甚至受控设备不能识别到对应的语音指令。而考虑到人耳的适应能力,人耳对环境噪音却并不敏感,人耳并不能准确感知环境噪音对语音识别的影响,因此,用户在不能识别的时候发出对应的控制指令将会造成受控设备不能识别或无法准确识别的问题,影响受控设备的操作。
上述内容仅用于辅助理解本发明的技术方案,并不代表承认上述内容是现有技术。
发明内容
本发明的主要目的旨在:使受控设备能够提示自身所识别出的环境噪声情况。
为实现上述目的,本发明实施例提供一种语音识别方法,包括以下步骤:
从环境中拾取出声音信息;
分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;
根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作。
优选地,所述根据对所述声音信息的分析结果,获取环境识别率对应的识别指数的步骤包括:
根据所述声音信息对应的噪声分贝,查找噪声识别策略表;
从所述噪声识别策略表中,获取所述噪声分贝映射的所述识别指数。
优选地,所述根据获取的所述识别指数,显示所述识别指数映射的提示信息的步骤包括:
根据获取的所述识别指数,查找所述识别指数对应的指示灯颜色;
按照找到的所述指示灯颜色,点亮对应的指示灯。
优选地,所述根据获取的所述识别指数,显示所述识别指数映射的提示信息的步骤包括:
根据获取的所述识别指数,查找所述识别指数对应的指示灯柱高度;
按照找到的所述指示灯柱高度,点亮对应高度的指示灯柱。
优选地,所述根据获取的所述识别指数,显示所述识别指数映射的提示信息的步骤包括:
根据获取的所述识别指数,查找所述识别指数对应的三原色值;
根据找出的所述三原色值,使对应的指示灯按照所述三原色值对应的颜色进行渐变显示。
优选地,所述分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数的步骤包括:
分析拾取出的所述声音信息,并根据对所述声音信息的分析结果,获取环境识别率对应的语音识别指数;
测量拾取所述声音信息的受控设备与发送语音指令控制所述受控设备的声源之间的距离,得到对应的距离参数,根据所述距离参数,获取所述距离参数对应的距离识别指数;
根据获取的所述语音识别指数和距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数,并将所述综合识别指数作为当前识别指数。
优选地,所述分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数的步骤包括:
分析拾取出的所述声音信息中各个预设频率的音频信号的强度或者比例;
根据预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,确定获取到的各个预设频率的音频信号的强度或者比例所对应的第一环境指数;
将所述环境指数作为当前识别指数。
优选地,所述将所述环境指数作为当前识别指数的步骤包括:
获取各个预设频率的信号中,中频信号的特征信息;
根据预设的特征信息与第二环境指数的映射关系,确定获取的所述特征信息所对应的第二环境指数,并按照预设的第一环境指数与第二环境指数的权重,对所述第一环境指数和第二环境指数进行加权平均得到当前环境指数;或者,根据预设的特征信息与环境指数偏量的映射关系,确定获取的所述特征信息所对应的环境指数偏量,采用所述环境偏量调整所述第一环境指数以得到当前识别指数。
优选地,所述获取各个预设频率的信号中,中频信号的特征信息的步骤替换为:
在获取到的第一环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,获取所述中频信号的特征信息。
此外,为实现上述目的,本发明还提出一种语音识别***,包括:
语音拾取模块,用于从环境中拾取出声音信息;
噪音分析模块,用于分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;
信息显示模块,用于根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作。
优选地,所述噪音分析模块还用于:
根据所述声音信息对应的噪声分贝,查找噪声识别策略表;
从所述噪声识别策略表中,获取所述噪声分贝映射的所述识别指数。
优选地,所述信息显示模块还用于:
根据获取的所述识别指数,查找所述识别指数对应的指示灯颜色;
按照找到的所述指示灯颜色,点亮对应的指示灯。
优选地,所述信息显示模块还用于:
根据获取的所述识别指数,查找所述识别指数对应的指示灯柱高度;
按照找到的所述指示灯柱高度,点亮对应高度的指示灯柱。
优选地,所述信息显示模块还用于:
根据获取的所述识别指数,查找所述识别指数对应的三原色值;
根据找出的所述三原色值,使对应的指示灯按照所述三原色值对应的颜色进行渐变显示。
优选地,所述噪音分析模块包括:
语音指数获取单元,用于分析拾取出的所述声音信息,并根据对所述声音信息的分析结果,获取环境识别率对应的语音识别指数;
距离指数获取单元,用于测量拾取所述声音信息的受控设备与发送语音指令控制所述受控设备的声源之间的距离,得到对应的距离参数,根据所述距离参数,获取所述距离参数对应的距离识别指数;
处理单元,用于根据获取的所述语音识别指数和距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数,并将所述综合识别指数作为当前识别指数。
优选地,所所述噪音分析模块包括:
分析单元,用于分析拾取出的所述声音信息中各个预设频率的音频信号的强度或者比例;
确定单元,用于根据预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,确定获取到的各个预设频率的音频信号的强度或者比例所对应的第一环境指数;
处理单元,用于将所述环境指数作为当前识别指数。
优选地,所述处理单元包括:
获取子单元,用于获取各个预设频率的信号中,中频信号的特征信息;
确定子单元,用于根据预设的特征信息与第二环境指数的映射关系,确定获取的所述特征信息所对应的第二环境指数,并按照预设的第一环境指数与第二环境指数的权重,对所述第一环境指数和第二环境指数进行加权平均得到当前环境指数;或者,根据预设的特征信息与环境指数偏量的映射关系,确定获取的所述特征信息所对应的环境指数偏量,采用所述环境偏量调整所述第一环境指数以得到当前识别指数。
优选地,所述获取子单元还用于在获取到的第一环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,获取所述中频信号的特征信息。
本发明实施例受控设备通过从环境中拾取出声音信息;分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;根据噪音分析模块获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作;使得所述受控设备能够将自身识别出的环境噪声情况及时显示出来,提示用户上述的环境噪声情况,便于用户根据所述受控设备提示的环境噪声情况,采取相应措施;提高了人机的可交互性。
附图说明
图1是本发明实现语音识别的受控设备较佳实施例的硬件架构示意图;
图2是本发明语音识别***较佳实施例的功能模块示意图;
图3是本发明语音识别方法较佳实施例的流程示意图。
本发明实施例目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
以下结合说明书附图及具体实施例进一步说明本发明的技术方案。应当理解,此处所描述的具体实施例仅仅用以解释本发明,并不用于限定本发明。
请参照图1,图1是本发明实现语音识别的受控设备较佳实施例的硬件架构示意图。如图1所示,所述受控设备100包括处理单元11、存储单元12、语音接收单元13、语音识别***14。其中,所述受控设备100包括但不限于:家用电器(比如空调器等)、移动终端(比如手机、平板电脑等)、遥控器或者其他任意适用的电子设备(优选空调器)。
所述存储单元12,用于存储该语音识别***14及其运行数据,以及该语音识别***14拾取的声音信息、与拾取的所述声音信息映射的环境识别率的识别指数等。需要强调的是,该存储单元12既可以是一个单独的存储装置,也可以是多个不同存储装置的统称,在此不作赘述。
所述语音接收单元13,用于在所述处理单元11的控制下,检测用户是否发送语音指令,并在检测到用户发送出语音指令后,拾取环境中的声音信息;或者,在所述处理单元11的控制下,实时或者按照预设周期拾取环境中的声音信息。
所述处理单元11,用于调用并执行所述语音识别***14,以从环境中拾取出声音信息;在拾取出对应的声音信息后,分析拾取出的所述声音信息,并根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;从而根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作,从而实现该受控设备100能够根据识别出的环境噪声情况来提示用户,供用户根据受控设备100提示的环境噪声情况来采取相应的处理策略。该处理单元11与存储单元12既可以分别是单独的单元,也可以集成在一起,构成一个控制器,在此不作赘述。
本实施例中,所述声音信息可以理解为用户发送语音控制指令所对应的声音信息;所述提示信息可以理解为:根据所述声音信息对应的环境识别率的识别指数的不同,所映射的不同的信息;比如,所述环境识别率对应的所述识别指数越高,表示环境中的噪音分别越低,则对应显示的提示信息也不同;比如,识别指数为高级时,显示绿色的指示灯;识别指数为中级时,显示黄色的指示灯;识别指数为低级时,显示红色的指示灯,从而提示用户发送对应语音指令时所需采取的对应策略。当然,所述提示信息也可以根据具体硬件设备、生产成本、使用环境等的不同,其具体表现形式也可以不同,在此不进行一一赘述。
请参照图2,图2是本发明语音识别***较佳实施例的功能模块示意图。
需要强调的是,对本领域的技术人员来说,图2所示功能模块图仅仅是一个较佳实施例的示例图,本领域的技术人员围绕图2所示的所述语音识别***14的功能模块,可轻易进行新的功能模块的补充;各功能模块的名称是自定义名称,仅用于辅助理解所述语音识别***14的各个程序功能块,不用于限定本发明的技术方案,本发明技术方案的核心是,各自定义名称的功能模块所要达成的功能。
本实施例提出的所述语音识别***14运行于图1所述受控设备100中,所述语音识别***14包括:语音拾取模块141、噪音分析模块142以及信息显示模块143,各个模块的功能如下:
语音拾取模块141,用于从环境中拾取出声音信息;
本实施例中,语音拾取模块141从当前环境中拾取的声音信息包括当前环境的噪声。
为了及时提示用户当前环境的噪声情况,便于用户根据所述语音识别***14提示的当前环境噪声情况,发出对应的语音指令,提高用户发送的所述语音指令的识别率,当语音识别***14运行的过程中,语音拾取模块141实时从当前环境中拾取出对应的声音信息。为了节约语音识别***14的运行成本、降低语音识别***14的功耗,语音拾取模块141也可以按照预设周期从当前环境中拾取出对应的声音信息。所述预设周期可由语音识别***14自定义比如设置所述预设周期为10ms,也可以由用户根据自身需求设置,或者在运行所述语音识别***14的受控设备100在出厂时,由技术人员根据所述语音识别***14的运行环境进行设置;本实施例对所述预设周期的具体时长不进行具体限制。当然,该预设周期也可以根据具体需要进行更新或者重新设置。
噪音分析模块142,用于分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;
当语音拾取模块141从环境中拾取出对应的声音信息时,噪音分析模块142分析语音拾取模块141所拾取出的声音信息,比如,获取所述声音信息中所包含的噪声对应的声音分贝,或者分析所拾取出的声音信息中是否包含有用户发送的语音指令。根据对上述声音信息的分析结果,所述噪音分析模块142获取当前环境识别率所对应的识别指数。
所述环境识别率对应的识别指数可以理解为:所述语音识别***14能够辨识用户发送的有效的语音指令的识别能力;所述识别指数越高,表示所述语音识别***14当前所处环境中的噪声越小,该语音识别***14对用户发送的语音指令的识别率越高。
信息显示模块143,用于根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作。
当噪音分析模块142根据对声音拾取模块141拾取的声音信息的分析结果,获取到环境识别率对应的识别指数后,所述信息显示模块143根据噪音分析模块142获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述信息显示模块143显示的所述提示信息执行相应操作。
比如,根据所述信息显示模块143显示的所述提示信息对应的当前环境的噪声为高级,则用户根据上述提示,选择是否发出相应的语音指令,或者用户是否需要高声发出相应的语音指令,或者用户是否需要降低当前环境中的噪声从而采取正常的声音分贝发出相应的语音指令等;再比如,根据所述信息显示模块143显示的所述提示信息对应的当前环境的噪声为低级,则表示当前环境下,环境噪声对所述语音识别***14识别语音指令的影响几乎可以忽略,即用户可以按照普通的方式发送相应的语音指令,则用户根据所述信息显示模块143的提示信息,按照普通方式发出相应的语音指令。
本发明实施例语音识别***的语音拾取模块从环境中拾取出声音信息;噪音分析模块分析语音拾取模块拾取出的所述声音信息,根据对所述声音信息的分析结果,噪音分析模块获取环境识别率对应的识别指数;信息显示模块根据噪音分析模块获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作;使得所述语音识别***能够将自身识别出的环境噪声情况及时显示出来,提示用户上述的环境噪声情况,便于用户根据所述语音识别***提示的环境噪声情况,采取相应措施;提高了语音识别***的智能性以及人机的可交互性;进一步地,由于用户能够根据提示出的环境噪声情况针对性地发送对应的语音指令,因此,提高了语音识别***对语音指令的识别准确率。
进一步地,请继续参照图1和图2,本发明语音识别***14中,所述噪音分析模块142根据对语音拾取模块141拾取的声音信息的分析结果,获取环境识别率对应的识别指数可以采用如下方式:
第一实施例:噪音分析模块142分析出语音拾取模块141拾取的声音信息所对应的噪声分贝时,噪音分析模块142查找噪声识别策略表,从所述噪声识别策略表中,所述噪音分析模块142获取所述噪声分贝映射的所述识别指数。进而,所述信息显示模块143根据所述噪音分析模块142获取的所述识别指数,显示对应的提示信息。
第二实施例:所述噪音分析模块142包括:
语音指数获取单元,用于分析拾取出的所述声音信息,并根据对所述声音信息的分析结果,获取环境识别率对应的语音识别指数;
当从环境中拾取出对应的声音信息时,语音指数获取单元分析所拾取出的声音信息,比如,获取所述声音信息中所包含的噪声对应的声音分贝,或者分析所拾取出的声音信息中是否包含有用户发送的语音指令。根据对上述声音信息的分析结果,所述语音指数获取单元获取当前环境识别率所对应的语音识别指数。
所述环境识别率对应的语音识别指数可以理解为:所述语音识别***14能够辨识用户发送的有效的语音指令的识别能力;所述语音识别指数越高,表示所述语音识别***14当前所处环境中的噪声越小,该语音识别***14对用户发送的语音指令的识别率越高。
距离指数获取单元,用于测量拾取所述声音信息的受控设备与发送语音指令控制所述受控设备的声源之间的距离,得到对应的距离参数,根据所述距离参数,获取所述距离参数对应的距离识别指数;
距离指数获取单元测量:拾取所述声音信息的受控设备即图1所述的受控设备100,与发送语音指令的所述声源之间的距离,得到对应的距离参数。在得到对应的距离参数后,所述距离指数获取单元测量根据上述距离参数,获取与所述距离参数映射的距离识别指数;所述距离识别指数可以理解为:表示发送有效语音指令的声源与所述受控设备100之间的距离远近,即所述声源与所述受控设备100的距离越近,则所述距离识别指数越高;所述声源与所述受控设备100的距离越远,则所述距离识别指数越低。而距离识别指数越高,则表示所述受控设备100对所述语音指令的识别率越高。
其中,所述距离指数获取单元测量所述用户与所述受控设备100之间的距离时,可以通过红外技术、距离传感器或者NFC技术等实现对上述距离的测量,本发明实施例对所述距离指数获取单元获取所述距离参数所采取的具体方式不做限定。另外,所述距离指数获取单元测量的所述用户与所述受控设备100之间的距离可以理解为:所述用户中心点与所述受控设备100之间的水平距离;通常情况下,可以设置所述距离指数获取单元获取的所述距离参数为发送语音指令的声源与所述受控设备100之间的水平距离;所述声源可理解为发送有效语音指令的用户、扬声器或者其他设备。本发明实施例中,优选的声源为用户。
处理单元,用于根据获取的所述语音识别指数和距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数,并将所述综合识别指数作为当前识别指数。
处理单元根据语音指数获取单元获取的语音识别指数,以及距离指数获取单元获取的距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数。所述预设规则可以理解为:受控设备100默认采用的计算规则,或者技术人员根据具体使用场景设置为所述受控设备100设置的计算规则;默认情况下,所述综合识别指数为语音识别指数与距离识别指数之和。比如,当前环境下,所述环境噪声对应的语音识别指数为52,对应的距离识别指数为-5,则语音识别指数与距离识别指数之和为47,即所述综合识别指数为47。
第三实施例:所述噪音分析模块142包括:
分析单元,用于分析拾取出的所述声音信息中各个预设频率的音频信号的强度或者比例;
在本实施例中,分析单元获取拾取出的声音信息中各个预设频率的音频信号的强度或者比例可通过以下方式实现:
a、通过不同频率的带通滤波器分别对取拾取出的声音信息进行过滤,以得到不同频率的音频信号,例如预设的频率分别为20HZ~300HZ、300
HZ~4KHZ以及4KHZ~20KHZ,通过带通滤波器可得到20HZ~300HZ、300
HZ~4KHZ以及4KHZ~20KHZ的第一音频信号、第二音频信号以及第三音频信号,则第一音频信号、第二音频信号以及第三音频信号的幅值a、b、c即为第一音频信号、第二音频信号以及第三音频信号的强度,而第一音频信号、第二音频信号以及第三音频信号的频率分别为a/(a+b+c)、b/(a+b+c)、c/(a+b+c)。
b、基于预设得到频率对取拾取出的声音信息进行傅里叶转换,以将该混合音频信号转换为对应频率频域信号,并基于侦测到的频域信号获取对应的各个频率信号的比例或者强度。
确定单元,用于根据预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,确定获取到的各个预设频率的音频信号的强度或者比例所对应的第一环境指数;
本实施例中,该预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,可为各个频率的音频信号的强度或者比例与环境指数之间的映射关系表,或者根据各个频率的音频信号的强度或者比例计算环境指数的公式,例如,预设的各个频率为低频、中频以及高频,例如低频信号、中频信号、以及高频信号的强度都较低,则对应的环境指数为30,而当低频信号和高频信号的强度较低、同时高频信号的强度较高,此时环境指数为45,该环境指数对应于环境的恶劣程度;而在其它实施例中,环境指数也可为当前环境清净程度与上述恶劣程度相反。
处理单元,用于将所述环境指数作为当前识别指数。
进一步地,为提高噪音提示的准确性,所述处理单元包括:获取子单元,用于获取各个预设频率的信号中,中频信号的特征信息;
确定子单元,用于根据预设的特征信息与第二环境指数的映射关系,确定获取的所述特征信息所对应的第二环境指数,并按照预设的第一环境指数与第二环境指数的权重,对所述第一环境指数和第二环境指数进行加权平均得到当前环境指数;或者,根据预设的特征信息与环境指数偏量的映射关系,确定获取的所述特征信息所对应的环境指数偏量,采用所述环境偏量调整所述第一环境指数以得到当前识别指数。
在本实施例中,该特征信息可包括脉冲信号、周期循环信号且强度基本不变以及无规律信号且强度大等三种。当中频信号的特征信息为脉冲时,说明该中频信号为人声的干扰,对识别极为不利,则第二环境指数较高如第二环境指数为80(该第二环境指数代表环境恶劣程度);当中频信号的特征信息为周期循环信号且强度基本不变时说明该中频信号为排风扇等固定设备发出的固定干扰,此类干扰容易滤除,对应第二环境指数中等如第二环境指数为65(该第二环境指数代表环境恶劣程度);当中频信号的特征信息为无规律信号且强度较大时说明该中频信号为的音源多且复杂,不易容易滤除,对应第二环境指数偏高如第二环境指数为90(该第二环境指数代表环境恶劣程度),生产厂商可预设特征信息与第二环境指数之间的映射关系,并根据获取到的特征信息调用对应的第二环境指数,该特征信息中脉冲信号的强度不同对应的第二环境指数也可不同,周期循环信号的强度不同对应的第二环境指数也可不同。
本领域技术人员可以理解的是,在得到第一环境指数后,可基于中频信号的特征信息得到对应的环境指数偏量,基于该环境指数偏量对所述第一环境指数进行调整,以得到当前环境指数。当中频信号的特征信息为脉冲时,说明该中频信号为人声的干扰,对识别极为不利,则对应环境指数偏量中等如环境指数偏量3;当中频信号的特征信息为周期循环信号且强度基本不变时说明该中频信号为排风扇等固定设备发出的固定干扰,此类干扰容易滤除,对应环境指数偏量较小如环境指数偏量-1;当中频信号的特征信息为无规律信号且强度较大时说明该中频信号为的音源多且复杂,不易容易滤除,对应环境指数偏量较大中等如环境指数偏量5,生产厂商可预设特征信息与环境指数偏量之间的映射关系,并根据获取到的特征信息调用对应的环境指数偏量,该特征信息中脉冲信号的强度不同对应的环境指数偏量也可不同,周期循环信号的强度不同对应的环境指数偏量也可不同。
进一步地,为减少***开销以节省能耗,所述获取子单元还用于在获取到的第一环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,获取所述中频信号的特征信息。
在本实施例中,在获取到的环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,第三获取模块才获取所述中频信号的特征信息,使得第二确定模块不用每次侦测到音频信号时均确定特征信息对应的第二环境指数或者环境指数偏量,节省***能耗。
在本实施例中,处理单元还用于在获取到的环境指数小于等于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度小于等于第二预设阀值时,将所述第一环境指数作为当前识别指数。
以上所列举出的三种获取环境识别率对应的识别指数的方式仅仅为示例性的,本领域技术人员利用本发明的技术思想,根据其具体需求所提出的其他获取环境识别率对应的识别指数的方式均在本发明的保护范围内,在此不进行一一穷举。
进一步地,本发明语音识别***实施例中,所述信息显示模块143根据所述噪音分析模块142获取的所述识别指数,显示所述识别指数映射的提示信息可以采用如下方式:
方式一:
利用指示灯的颜色来显示不同分贝的噪声,比如,所述噪音分析模块142获取的所述噪声分贝为50db(分贝)以下,则所述信息显示模块143显示绿灯;所述噪音分析模块142获取的所述噪声分贝为50db~70db(包含50db,不包含70db)之间,则所述信息显示模块143显示黄灯;所述噪音分析模块142获取的所述噪声分贝为70db以上(含70db),则所述信息显示模块143显示红灯。比如,若噪音分析模块142分析出语音拾取模块141拾取的声音信息所对应的噪声分贝为45db,则所述信息显示模块143显示绿灯。
方式二:
利用指示灯柱的高度来显示不同分贝的噪声,其中,所述指示灯柱可以由若干个显示单元水平或垂直排列,并根据不同的噪声分贝计算需要显示的显示单元的个数,从而对应显示指示灯柱的高度。比如,将10个显示单元组成的指示灯柱垂直排列,当所述噪音分析模块142获取的所述噪声分贝为40db时,所述信息显示模块143点亮组成指示灯柱最下方的一个显示单元;当所述噪音分析模块142获取的所述噪声分贝为70db时,所述信息显示模块143点亮从组成指示灯柱的最下到最上方的所有10个显示单元,由显示单元的个数对应指示灯柱的高度来提醒用户环境的噪音水平。再比如,所述噪音分析模块142获取的当前环境的噪声分贝为65db,则所述信息显示模块143点亮8个显示单元,此时用户看到的灯柱比较高,则知道需要用比较大的声音,以便于所述语音识别***14能够准确识别对应的语音指令。
方式三:
利用指示灯的三原色值来显示不同分贝的噪声,将环境中噪声分贝值的大小变化转换成对应的RGB三原色的成分,通过根据噪声分贝值来调整三原色中各原色成分的亮度值,从而配比出渐变色,随着噪声分贝值的大小进行显示颜色的渐变。比如,外界噪音变大时,相应地所述噪音分析模块142获取的所述噪声分贝值也会变大,则所述信息显示模块143根据所述噪声分贝值调配出来的颜色值也由淡色系逐渐往深色系变换,从而提示用户噪声越来越大。
方式四:
利用指示灯的闪烁频率来显示不同分贝的噪声,比如当所述噪音分析模块142获取的所述噪声分贝值在第一预设分贝值以下时,则所述信息显示模块143显示的指示灯为常亮,也就是说,人眼看到的所述指示灯不进行闪烁;随着噪声越来越大,则所述信息显示模块143显示的指示灯的闪烁频率也越来越高,从而提示用户噪声越来越大。也可以理解为,所述信息显示模块143显示的指示灯的闪烁频率与环境噪声的大小成正比。
以上所列举出的四种显示方式仅仅为示例性的,本领域技术人员利用本发明的技术思想,根据其具体需求所提出的其他显示方式均在本发明的保护范围内,在此不进行一一穷举。
本发明语音识别***通过将识别的噪声显示出来以提示用户当前环境中的噪声情况,提高了该语音识别***的智能性;进一步地,由于能够根据提示出的环境噪声情况针对性地发送对应的语音指令,因此,提高了语音识别***对语音指令的识别准确率。
请参照图3,图3是本发明语音识别方法较佳实施例的流程示意图,本发明实施例提供了一种语音识别方法一实施例。
如图3所示,本发明语音识别方法包括以下步骤:
步骤S01、从环境中拾取出声音信息;
本实施例中,受控设备从当前环境中拾取的声音信息包括当前环境的噪声。
为了及时提示用户当前环境的噪声情况,便于用户根据所述受控设备提示的当前环境噪声情况,发出对应的语音指令,提高用户发送的所述语音指令的识别率,当受控设备运行的过程中,受控设备实时从当前环境中拾取出对应的声音信息。为了节约受控设备的运行成本、降低受控设备的功耗,受控设备也可以按照预设周期从当前环境中拾取出对应的声音信息。所述预设周期可由受控设备自定义比如设置所述预设周期为10ms,也可以由用户根据自身需求设置,或者在受控设备在出厂时,由技术人员根据所述受控设备的运行环境进行设置;本实施例对所述预设周期的具体时长不进行具体限制。当然,该预设周期也可以根据具体需要进行更新或者重新设置。
步骤S02、分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;
当从环境中拾取出对应的声音信息时,受控设备分析所拾取出的声音信息,比如,获取所述声音信息中所包含的噪声对应的声音分贝,或者分析所拾取出的声音信息中是否包含有用户发送的语音指令。根据对上述声音信息的分析结果,所述受控设备获取当前环境识别率所对应的识别指数。
所述环境识别率对应的识别指数可以理解为:所述受控设备能够辨识用户发送的有效的语音指令的识别能力;所述识别指数越高,表示所述受控设备当前所处环境中的噪声越小,该受控设备对用户发送的语音指令的识别率越高。
步骤S03、根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作。
当受控设备根据拾取的声音信息的分析结果,获取到环境识别率对应的识别指数后,受控设备根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述受控设备显示的所述提示信息执行相应操作。
比如,根据所述受控设备显示的所述提示信息对应的当前环境的噪声为高级,则用户根据上述提示,选择是否发出相应的语音指令,或者用户是否需要高声发出相应的语音指令,或者用户是否需要降低当前环境中的噪声从而采取正常的声音分贝发出相应的语音指令等;再比如,根据所述受控设备显示的所述提示信息对应的当前环境的噪声为低级,则表示当前环境下,环境噪声对所述受控设备识别语音指令的影响几乎可以忽略,即用户可以按照普通的方式发送相应的语音指令,则用户根据所述受控设备的提示信息,按照普通方式发出相应的语音指令。
本发明实施例受控设备从环境中拾取出声音信息;分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;根据噪音分析模块获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作;使得所述受控设备能够将自身识别出的环境噪声情况及时显示出来,提示用户上述的环境噪声情况,便于用户根据所述受控设备提示的环境噪声情况,采取相应措施;提高了人机的可交互性;进一步地,由于用户能够根据提示出的环境噪声情况针对性地发送对应的语音指令,因此,提高了受控设备对语音指令的识别准确率。
进一步地,请继续参照图3,本发明实施例语音识别方法中,受控设备根据对所述声音信息的分析结果,获取环境识别率对应的识别指数可以采用如下方式:
第一实施例:受控设备分析出拾取的声音信息所对应的噪声分贝时,查找噪声识别策略表,从所述噪声识别策略表中,获取所述噪声分贝映射的所述识别指数。进而,根据获取的所述识别指数,受控设备显示对应的提示信息。
第二实施例:所述步骤S02包括:
分析拾取出的所述声音信息,并根据对所述声音信息的分析结果,获取环境识别率对应的语音识别指数;
测量拾取所述声音信息的受控设备与发送语音指令控制所述受控设备的声源之间的距离,得到对应的距离参数,根据所述距离参数,获取所述距离参数对应的距离识别指数;
根据获取的所述语音识别指数和距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数,并将所述综合识别指数作为当前识别指数。
当从环境中拾取出对应的声音信息时,受控设备分析所拾取出的声音信息,比如,获取所述声音信息中所包含的噪声对应的声音分贝,或者分析所拾取出的声音信息中是否包含有用户发送的语音指令。根据对上述声音信息的分析结果,所述受控设备获取当前环境识别率所对应的语音识别指数。
所述环境识别率对应的语音识别指数可以理解为:所述语音识别***14能够辨识用户发送的有效的语音指令的识别能力;所述语音识别指数越高,表示所述语音识别***14当前所处环境中的噪声越小,该语音识别***14对用户发送的语音指令的识别率越高。
受控设备获取其与发送语音指令的所述声源之间的距离,得到对应的距离参数。在得到对应的距离参数后,所述受控设备测量根据上述距离参数,获取与所述距离参数映射的距离识别指数;所述距离识别指数可以理解为:表示发送有效语音指令的声源与所述受控设备之间的距离远近,即所述声源与所述受控设备的距离越近,则所述距离识别指数越高;所述声源与所述受控设备的距离越远,则所述距离识别指数越低。而距离识别指数越高,则表示所述受控设备对所述语音指令的识别率越高。
其中,所述受控设备测量其与所述用户之间的距离时,可以通过红外技术、距离传感器或者NFC技术等实现对上述距离的测量,本发明实施例对所述受控设备获取所述距离参数所采取的具体方式不做限定。另外,所述受控设备其与所述用户之间的距离可以理解为:所述用户中心点与所述受控设备之间的水平距离;通常情况下,可以设置所述距离参数为发送语音指令的声源与所述受控设备之间的水平距离;所述声源可理解为发送有效语音指令的用户、扬声器或者其他设备。本发明实施例中,优选的声源为用户。
处理单元,用于根据获取的所述语音识别指数和距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数,并将所述综合识别指数作为当前识别指数。
被控终端根据语音识别指数,以及距离指数获取单元获取的距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数。所述预设规则可以理解为:受控设备默认采用的计算规则,或者技术人员根据具体使用场景设置为所述受控设备设置的计算规则;默认情况下,所述综合识别指数为语音识别指数与距离识别指数之和。比如,当前环境下,所述环境噪声对应的语音识别指数为52,对应的距离识别指数为-5,则语音识别指数与距离识别指数之和为47,即所述综合识别指数为47。
第三实施例,所述步骤S20包括:
分析拾取出的所述声音信息中各个预设频率的音频信号的强度或者比例;
根据预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,确定获取到的各个预设频率的音频信号的强度或者比例所对应的第一环境指数;
将所述环境指数作为当前识别指数。
在本实施例中,获取侦测到的混合音频信号中各个预设频率的音频信号的强度或者比例可通过以下方式实现:
a、通过不同频率的带通滤波器分别对侦测到的混合音频信号进行过滤,以得到不同频率的音频信号,例如预设的频率分别为20HZ~300HZ、300
HZ~4KHZ以及4KHZ~20KHZ,通过带通滤波器可得到20HZ~300HZ、300
HZ~4KHZ以及4KHZ~20KHZ的第一音频信号、第二音频信号以及第三音频信号,则第一音频信号、第二音频信号以及第三音频信号的幅值a、b、c即为第一音频信号、第二音频信号以及第三音频信号的强度,而第一音频信号、第二音频信号以及第三音频信号的频率分别为a/(a+b+c)、b/(a+b+c)、c/(a+b+c)。
b、基于预设得到频率对侦测到的混合音频信号进行傅里叶转换,以将该混合音频信号转换为对应频率频域信号,并基于侦测到的频域信号获取对应的各个频率信号的比例或者强度。
在本实施例中,该预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,可为各个频率的音频信号的强度或者比例与环境指数之间的映射关系表,或者根据各个频率的音频信号的强度或者比例计算环境指数的公式,例如,预设的各个频率为低频、中频以及高频,例如低频信号、中频信号、以及高频信号的强度都较低,则对应的环境指数为30,而当低频信号和高频信号的强度较低、同时高频信号的强度较高,此时环境指数为45,该环境指数对应于环境的恶劣程度;而在其它实施例中,环境指数也可为当前环境清净程度与上述恶劣程度相反。
进一步地,为提高噪音提示的准确性,所述将所述环境指数作为当前识别指数的步骤包括:
获取各个预设频率的信号中,中频信号的特征信息;
根据预设的特征信息与第二环境指数的映射关系,确定获取的所述特征信息所对应的第二环境指数,并按照预设的第一环境指数与第二环境指数的权重,对所述第一环境指数和第二环境指数进行加权平均得到当前环境指数;或者,根据预设的特征信息与环境指数偏量的映射关系,确定获取的所述特征信息所对应的环境指数偏量,采用所述环境偏量调整所述第一环境指数以得到当前识别指数。
在本实施例中,该特征信息可包括脉冲信号、周期循环信号且强度基本不变以及无规律信号且强度大等三种。当中频信号的特征信息为脉冲时,说明该中频信号为人声的干扰,对识别极为不利,则第二环境指数较高如第二环境指数为80(该第二环境指数代表环境恶劣程度);当中频信号的特征信息为周期循环信号且强度基本不变时说明该中频信号为排风扇等固定设备发出的固定干扰,此类干扰容易滤除,对应第二环境指数中等如第二环境指数为65(该第二环境指数代表环境恶劣程度);当中频信号的特征信息为无规律信号且强度较大时说明该中频信号为的音源多且复杂,不易容易滤除,对应第二环境指数偏高如第二环境指数为90(该第二环境指数代表环境恶劣程度),生产厂商可预设特征信息与第二环境指数之间的映射关系,并根据获取到的特征信息调用对应的第二环境指数,该特征信息中脉冲信号的强度不同对应的第二环境指数也可不同,周期循环信号的强度不同对应的第二环境指数也可不同。
本领域技术人员可以理解的是,在得到第一环境指数后,可基于中频信号的特征信息得到对应的环境指数偏量,基于该环境指数偏量对所述第一环境指数进行调整,以得到当前环境指数。当中频信号的特征信息为脉冲时,说明该中频信号为人声的干扰,对识别极为不利,则对应环境指数偏量中等如环境指数偏量3;当中频信号的特征信息为周期循环信号且强度基本不变时说明该中频信号为排风扇等固定设备发出的固定干扰,此类干扰容易滤除,对应环境指数偏量较小如环境指数偏量-1;当中频信号的特征信息为无规律信号且强度较大时说明该中频信号为的音源多且复杂,不易容易滤除,对应环境指数偏量较大中等如环境指数偏量5,生产厂商可预设特征信息与环境指数偏量之间的映射关系,并根据获取到的特征信息调用对应的环境指数偏量,该特征信息中脉冲信号的强度不同对应的环境指数偏量也可不同,周期循环信号的强度不同对应的环境指数偏量也可不同。
进一步地,为减少***开销以节省能耗,所述获取各个预设频率的信号中,中频信号的特征信息的步骤替换为:
在获取到的第一环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,获取所述中频信号的特征信息。
在本实施例中,在获取到的环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,第三获取模块才获取所述中频信号的特征信息,使得第二确定模块不用每次侦测到音频信号时均确定特征信息对应的第二环境指数或者环境指数偏量,节省***能耗。
在本实施例中,处理单元还用于在获取到的环境指数小于等于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度小于等于第二预设阀值时,将所述第一环境指数作为当前识别指数。
以上所列举出的三种获取环境识别率对应的识别指数的方式仅仅为示例性的,本领域技术人员利用本发明的技术思想,根据其具体需求所提出的其他获取环境识别率对应的识别指数的方式均在本发明的保护范围内,在此不进行一一穷举。
进一步地,受控设备根据获取的所述识别指数,显示所述识别指数映射的提示信息可以采用如下方式:
方式一:
受控设备利用指示灯的颜色来显示不同分贝的噪声,比如,所述受控设备获取的所述噪声分贝为50db(分贝)以下,则所述受控设备显示绿灯;所述受控设备获取的所述噪声分贝为50db~70db(包含50db,不包含70db)之间,则所述受控设备显示黄灯;所述受控设备获取的所述噪声分贝为70db以上(含70db),则所述受控设备显示红灯。比如,若受控设备分析出语音拾取模块141拾取的声音信息所对应的噪声分贝为45db,则所述受控设备显示绿灯。
方式二:
受控设备利用指示灯柱的高度来显示不同分贝的噪声,其中,所述指示灯柱可以由若干个显示单元水平或垂直排列,并根据不同的噪声分贝计算需要显示的显示单元的个数,从而对应显示指示灯柱的高度。比如,将10个显示单元组成的指示灯柱垂直排列,当所述受控设备获取的所述噪声分贝为40db时,所述受控设备点亮组成指示灯柱最下方的一个显示单元;当所述受控设备获取的所述噪声分贝为70db时,所述受控设备点亮从组成指示灯柱的最下到最上方的所有10个显示单元,由显示单元的个数对应指示灯柱的高度来提醒用户环境的噪音水平。再比如,所述受控设备获取的当前环境的噪声分贝为65db,则所述受控设备点亮8个显示单元,此时用户看到的灯柱比较高,则知道需要用比较大的声音,以便于所述受控设备能够准确识别对应的语音指令。
方式三:
利用指示灯的三原色值来显示不同分贝的噪声,将环境中噪声分贝值的大小变化转换成对应的RGB三原色的成分,通过根据噪声分贝值来调整三原色中各原色成分的亮度值,从而配比出渐变色,随着噪声分贝值的大小进行显示颜色的渐变。比如,外界噪音变大时,相应地所述受控设备获取的所述噪声分贝值也会变大,则所述受控设备根据所述噪声分贝值调配出来的颜色值也由淡色系逐渐往深色系变换,从而提示用户噪声越来越大。
方式四:
利用指示灯的闪烁频率来显示不同分贝的噪声,比如当所述受控设备获取的所述噪声分贝值在第一预设分贝值以下时,则所述受控设备显示的指示灯为常亮,也就是说,人眼看到的所述指示灯不进行闪烁;随着噪声越来越大,则所述受控设备显示的指示灯的闪烁频率也越来越高,从而提示用户噪声越来越大。也可以理解为,所述受控设备显示的指示灯的闪烁频率与环境噪声的大小成正比。
以上所列举出的四种显示方式仅仅为示例性的,本领域技术人员利用本发明的技术思想,根据其具体需求所提出的其他显示方式均在本发明的保护范围内,在此不进行一一穷举。
本发明受控设备通过将识别的噪声显示出来以提示用户当前环境中的噪声情况,提高了受控设备的智能性;进一步地,由于能够根据提示出的环境噪声情况针对性地发送对应的语音指令,因此,提高了受控设备对语音指令的识别准确率。
需要说明的是,在本文中,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者***不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者***所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括该要素的过程、方法、物品或者***中还存在另外的相同要素。
上述本发明实施例序号仅仅为了描述,不代表实施例的优劣。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。基于这样的理解,本发明的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中,包括若干指令用以使得一台终端设备(可以是手机,计算机,服务器,空调器,或者网络设备等)执行本发明各个实施例所述的方法。
以上所述仅为本发明的优选实施例,并非因此限制其专利范围,凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换,直接或间接运用在其他相关的技术领域,均同理包括在本发明的专利保护范围内。
Claims (18)
- 一种语音识别方法,其特征在于,包括以下步骤:从环境中拾取出声音信息;分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作。
- 如权利要求1所述语音识别方法,其特征在于,所述根据对所述声音信息的分析结果,获取环境识别率对应的识别指数的步骤包括:根据所述声音信息对应的噪声分贝,查找噪声识别策略表;从所述噪声识别策略表中,获取所述噪声分贝映射的所述识别指数。
- 如权利要求1或2所述的语音识别方法,其特征在于,所述根据获取的所述识别指数,显示所述识别指数映射的提示信息的步骤包括:根据获取的所述识别指数,查找所述识别指数对应的指示灯颜色;按照找到的所述指示灯颜色,点亮对应的指示灯。
- 如权利要求1或2所述的语音识别方法,其特征在于,所述根据获取的所述识别指数,显示所述识别指数映射的提示信息的步骤包括:根据获取的所述识别指数,查找所述识别指数对应的指示灯柱高度;按照找到的所述指示灯柱高度,点亮对应高度的指示灯柱。
- 如权利要求1或2所述的语音识别方法,其特征在于,所述根据获取的所述识别指数,显示所述识别指数映射的提示信息的步骤包括:根据获取的所述识别指数,查找所述识别指数对应的三原色值;根据找出的所述三原色值,使对应的指示灯按照所述三原色值对应的颜色进行渐变显示。
- 如权利要求1所述的语音识别方法,其特征在于,所述分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数的步骤包括:分析拾取出的所述声音信息,并根据对所述声音信息的分析结果,获取环境识别率对应的语音识别指数;测量拾取所述声音信息的受控设备与发送语音指令控制所述受控设备的声源之间的距离,得到对应的距离参数,根据所述距离参数,获取所述距离参数对应的距离识别指数;根据获取的所述语音识别指数和距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数,并将所述综合识别指数作为当前识别指数。
- 如权利要求1所述的语音识别方法,其特征在于,所述分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数的步骤包括:分析拾取出的所述声音信息中各个预设频率的音频信号的强度或者比例;根据预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,确定获取到的各个预设频率的音频信号的强度或者比例所对应的第一环境指数;将所述环境指数作为当前识别指数。
- 如权利要求7所述的语音识别方法,其特征在于,所述将所述环境指数作为当前识别指数的步骤包括:获取各个预设频率的信号中,中频信号的特征信息;根据预设的特征信息与第二环境指数的映射关系,确定获取的所述特征信息所对应的第二环境指数,并按照预设的第一环境指数与第二环境指数的权重,对所述第一环境指数和第二环境指数进行加权平均得到当前环境指数;或者,根据预设的特征信息与环境指数偏量的映射关系,确定获取的所述特征信息所对应的环境指数偏量,采用所述环境偏量调整所述第一环境指数以得到当前识别指数。
- 如权利要求8所述的语音识别方法,其特征在于,所述获取各个预设频率的信号中,中频信号的特征信息的步骤替换为:在获取到的第一环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,获取所述中频信号的特征信息。
- 一种语音识别***,其特征在于,包括:语音拾取模块,用于从环境中拾取出声音信息;噪音分析模块,用于分析拾取出的所述声音信息,根据对所述声音信息的分析结果,获取环境识别率对应的识别指数;信息显示模块,用于根据获取的所述识别指数,显示所述识别指数映射的提示信息,供用户根据所述提示信息执行相应操作。
- 如权利要求10所述的语音识别***,其特征在于,所述噪音分析模块还用于:根据所述声音信息对应的噪声分贝,查找噪声识别策略表;从所述噪声识别策略表中,获取所述噪声分贝映射的所述识别指数。
- 如权利要求10或11所述的语音识别***,其特征在于,所述信息显示模块还用于:根据获取的所述识别指数,查找所述识别指数对应的指示灯颜色;按照找到的所述指示灯颜色,点亮对应的指示灯。
- 如权利要求10或11所述的语音识别***,其特征在于,所述信息显示模块还用于:根据获取的所述识别指数,查找所述识别指数对应的指示灯柱高度;按照找到的所述指示灯柱高度,点亮对应高度的指示灯柱。
- 如权利要求10或11所述的语音识别***,其特征在于,所述信息显示模块还用于:根据获取的所述识别指数,查找所述识别指数对应的三原色值;根据找出的所述三原色值,使对应的指示灯按照所述三原色值对应的颜色进行渐变显示。
- 如权利要求10所述的语音识别***,其特征在于,所述噪音分析模块包括:语音指数获取单元,用于分析拾取出的所述声音信息,并根据对所述声音信息的分析结果,获取环境识别率对应的语音识别指数;距离指数获取单元,用于测量拾取所述声音信息的受控设备与发送语音指令控制所述受控设备的声源之间的距离,得到对应的距离参数,根据所述距离参数,获取所述距离参数对应的距离识别指数;处理单元,用于根据获取的所述语音识别指数和距离识别指数,按照预设规则计算获取环境识别率对应的综合识别指数,并将所述综合识别指数作为当前识别指数。
- 如权利要求10所述的语音识别***,其特征在于,所述噪音分析模块包括:分析单元,用于分析拾取出的所述声音信息中各个预设频率的音频信号的强度或者比例;确定单元,用于根据预设的各个频率的音频信号的强度或者比例与环境指数之间的映射关系,确定获取到的各个预设频率的音频信号的强度或者比例所对应的第一环境指数;处理单元,用于将所述环境指数作为当前识别指数。
- 如权利要求16所述的语音识别***,其特征在于,所述处理单元包括:获取子单元,用于获取各个预设频率的信号中,中频信号的特征信息;确定子单元,用于根据预设的特征信息与第二环境指数的映射关系,确定获取的所述特征信息所对应的第二环境指数,并按照预设的第一环境指数与第二环境指数的权重,对所述第一环境指数和第二环境指数进行加权平均得到当前环境指数;或者,根据预设的特征信息与环境指数偏量的映射关系,确定获取的所述特征信息所对应的环境指数偏量,采用所述环境偏量调整所述第一环境指数以得到当前识别指数。
- 如权利要求17所述的语音识别方法,其特征在于,所述获取子单元还用于在获取到的第一环境指数大于第一预设阀值时,或者各个预设频率的信号中,中频信号的频率或强度大于第二预设阀值时,获取所述中频信号的特征信息。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410300797.7A CN104078040A (zh) | 2014-06-26 | 2014-06-26 | 语音识别方法及*** |
CN201410300797.7 | 2014-06-26 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2015196720A1 true WO2015196720A1 (zh) | 2015-12-30 |
Family
ID=51599263
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2014/091947 WO2015196720A1 (zh) | 2014-06-26 | 2014-11-21 | 语音识别方法及*** |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN104078040A (zh) |
WO (1) | WO2015196720A1 (zh) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111508473A (zh) * | 2020-06-12 | 2020-08-07 | 佛山科学技术学院 | 一种语音识别率分析装置、方法及存储介质 |
CN112015276A (zh) * | 2020-09-09 | 2020-12-01 | 深圳市南方亿信计算机信息***有限公司 | 一种包含显示效果调节和音量控制的人机交互方法 |
CN112327789A (zh) * | 2020-11-26 | 2021-02-05 | 江西台德智慧科技有限公司 | 一种应用于智能语音助手的语音交互的***及方法 |
CN114401566A (zh) * | 2022-01-19 | 2022-04-26 | 深圳市佳乐鑫科技有限公司 | 一种基于物联网的智能家居led小夜灯 |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104078040A (zh) * | 2014-06-26 | 2014-10-01 | 美的集团股份有限公司 | 语音识别方法及*** |
CN105100352B (zh) * | 2015-06-24 | 2018-09-25 | 小米科技有限责任公司 | 获取联系人信息的方法及装置 |
CN105355201A (zh) * | 2015-11-27 | 2016-02-24 | 百度在线网络技术(北京)有限公司 | 基于场景的语音服务处理方法、装置和终端设备 |
CN106796790B (zh) * | 2016-11-16 | 2020-11-10 | 深圳达闼科技控股有限公司 | 机器人语音指令识别的方法及相关机器人装置 |
CN108652633A (zh) * | 2018-05-03 | 2018-10-16 | 华南理工大学广州学院 | 一种炫彩肺活量测试仪 |
CN111380189B (zh) * | 2020-03-27 | 2021-05-04 | 四川虹美智能科技有限公司 | 智能空调控制方法及智能空调 |
CN112208686A (zh) * | 2020-09-16 | 2021-01-12 | 纳恩博(天津)科技有限公司 | 数据处理方法及电动摩托车 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007212704A (ja) * | 2006-02-09 | 2007-08-23 | Univ Waseda | 雑音スペクトル推定方法、雑音抑圧方法及び雑音抑圧装置 |
US20090012786A1 (en) * | 2007-07-06 | 2009-01-08 | Texas Instruments Incorporated | Adaptive Noise Cancellation |
CN102907019A (zh) * | 2010-07-29 | 2013-01-30 | 英派尔科技开发有限公司 | 通过控制电子设备操作的声学噪声管理 |
CN103065631A (zh) * | 2013-01-24 | 2013-04-24 | 华为终端有限公司 | 一种语音识别的方法、装置 |
CN103677234A (zh) * | 2012-09-17 | 2014-03-26 | 联想(北京)有限公司 | 一种信息处理的方法及电子设备 |
CN104078041A (zh) * | 2014-06-26 | 2014-10-01 | 美的集团股份有限公司 | 语音识别方法及*** |
CN104078040A (zh) * | 2014-06-26 | 2014-10-01 | 美的集团股份有限公司 | 语音识别方法及*** |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1965218A (zh) * | 2004-06-04 | 2007-05-16 | 皇家飞利浦电子股份有限公司 | 交互式语音识别***的性能预测 |
CN1773847A (zh) * | 2004-11-12 | 2006-05-17 | 乐金电子(昆山)电脑有限公司 | 便携式终端机的音量自动调节装置以及方法 |
CN1897054A (zh) * | 2005-07-14 | 2007-01-17 | 松下电器产业株式会社 | 可根据声音种类发出警报的传输装置及方法 |
CN101790000B (zh) * | 2010-02-20 | 2014-08-13 | 华为终端有限公司 | 一种环境声音提醒方法和移动终端 |
CN103594088A (zh) * | 2013-11-11 | 2014-02-19 | 联想(北京)有限公司 | 一种信息处理方法和电子设备 |
-
2014
- 2014-06-26 CN CN201410300797.7A patent/CN104078040A/zh active Pending
- 2014-11-21 WO PCT/CN2014/091947 patent/WO2015196720A1/zh active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007212704A (ja) * | 2006-02-09 | 2007-08-23 | Univ Waseda | 雑音スペクトル推定方法、雑音抑圧方法及び雑音抑圧装置 |
US20090012786A1 (en) * | 2007-07-06 | 2009-01-08 | Texas Instruments Incorporated | Adaptive Noise Cancellation |
CN102907019A (zh) * | 2010-07-29 | 2013-01-30 | 英派尔科技开发有限公司 | 通过控制电子设备操作的声学噪声管理 |
CN103677234A (zh) * | 2012-09-17 | 2014-03-26 | 联想(北京)有限公司 | 一种信息处理的方法及电子设备 |
CN103065631A (zh) * | 2013-01-24 | 2013-04-24 | 华为终端有限公司 | 一种语音识别的方法、装置 |
CN104078041A (zh) * | 2014-06-26 | 2014-10-01 | 美的集团股份有限公司 | 语音识别方法及*** |
CN104078040A (zh) * | 2014-06-26 | 2014-10-01 | 美的集团股份有限公司 | 语音识别方法及*** |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111508473A (zh) * | 2020-06-12 | 2020-08-07 | 佛山科学技术学院 | 一种语音识别率分析装置、方法及存储介质 |
CN111508473B (zh) * | 2020-06-12 | 2023-03-14 | 佛山科学技术学院 | 一种语音识别率分析装置、方法及存储介质 |
CN112015276A (zh) * | 2020-09-09 | 2020-12-01 | 深圳市南方亿信计算机信息***有限公司 | 一种包含显示效果调节和音量控制的人机交互方法 |
CN112015276B (zh) * | 2020-09-09 | 2023-06-02 | 深圳华创电科技术有限公司 | 一种包含显示效果调节和音量控制的人机交互方法 |
CN112327789A (zh) * | 2020-11-26 | 2021-02-05 | 江西台德智慧科技有限公司 | 一种应用于智能语音助手的语音交互的***及方法 |
CN112327789B (zh) * | 2020-11-26 | 2023-04-28 | 江西台德智慧科技有限公司 | 一种应用于智能语音助手的语音交互的***及方法 |
CN114401566A (zh) * | 2022-01-19 | 2022-04-26 | 深圳市佳乐鑫科技有限公司 | 一种基于物联网的智能家居led小夜灯 |
CN114401566B (zh) * | 2022-01-19 | 2024-04-12 | 深圳市佳乐鑫科技有限公司 | 一种基于物联网的智能家居led小夜灯 |
Also Published As
Publication number | Publication date |
---|---|
CN104078040A (zh) | 2014-10-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2015196720A1 (zh) | 语音识别方法及*** | |
WO2019051899A1 (zh) | 终端控制方法、装置及存储介质 | |
WO2020246844A1 (en) | Device control method, conflict processing method, corresponding apparatus and electronic device | |
WO2020045950A1 (en) | Method, device, and system of selectively using multiple voice data receiving devices for intelligent service | |
WO2016104988A9 (ko) | 이동 단말기, 오디오 출력장치, 및 이를 구비하는 오디오 출력 시스템 | |
WO2015158132A1 (zh) | 语音控制方法和*** | |
WO2015037963A1 (en) | Home appliance and mobile terminal | |
WO2014062032A1 (en) | Display device, remote control device to control display device, method of controlling display device, method of controlling server and method of controlling remote control device | |
WO2019051908A1 (zh) | 终端控制方法、装置及计算机可读存储介质 | |
WO2020130535A1 (en) | Electronic device including earphone, and method of controlling the electronic device | |
WO2020060368A2 (en) | Method and apparatus for providing notification by interworking plurality of electronic devices | |
WO2016167552A1 (ko) | 인체 착용 장치 및 이의 동작 방법 | |
WO2019037319A1 (zh) | 电量预警方法、服务器、移动终端及存储介质 | |
WO2016204496A1 (en) | System and method of providing information of peripheral device | |
WO2020231202A1 (ko) | 복수 개의 스피커들을 포함하는 전자 장치 및 그 제어 방법 | |
WO2019051895A1 (zh) | 终端控制方法、装置及存储介质 | |
WO2019051890A1 (zh) | 终端控制方法、装置及计算机可读存储介质 | |
WO2016029597A1 (zh) | 终端控制方法和*** | |
WO2019071762A1 (zh) | 楼层位置定位方法、***、服务器和计算机可读存储介质 | |
WO2019172565A1 (ko) | 외부 장치와의 연결을 위한 전자 장치 및 방법 | |
WO2015158133A1 (zh) | 语音控制指令纠错方法和*** | |
WO2017088437A1 (zh) | 智能电视控制终端摄像头切换方法及装置 | |
WO2019051934A1 (zh) | 业务人员考核方法、考核平台和计算机可读存储介质 | |
WO2019042220A1 (zh) | 智能***调整方法、装置及计算机可读存储介质 | |
WO2014109504A1 (ko) | 거리 검출 장치, 및 이를 구비하는 영상처리장치 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14896153 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14896153 Country of ref document: EP Kind code of ref document: A1 |