• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    PURPOSE: A speech recognition method of a mobile communication terminal is provided to reduce power consumed for speech recognition of the mobile terminal. CONSTITUTION: It is judged whether a user pushes a speech recognition key(S11). When it is judged that the user pushes the speech recognition key, a signal input through a microphone and an analog/digital converter is stored in a memory(S12). It is judged whether the speech recognition key is returned to the initial state(S13). When it is judged that the speech recognition key is returned, it is decided whether the operation of storing data in the memory(S14). A characteristic parameter is extracted from the data stored in the memory to recognize the speech of the user through a speech recognition engine(S15,S16).
    公开号:KR20040042942A
    申请号:KR1020020070894
    申请日:2002-11-14
    公开日:2004-05-22
    发明作者:이상헌
    申请人:엘지전자 주식회사；
    IPC主号:
    专利说明:

    VOICE RECOGNITION METHOD FOR MOBILE COMMUNICATION TERMINAL}
    [2] The present invention relates to a voice recognition method of a mobile communication terminal, and more particularly, to a voice recognition method of a mobile communication terminal that can reduce the power required for voice recognition in a portable terminal, the life of the battery is sensitive.
    [3] In general, speech recognition technology is a state-of-the-art software technology that extracts and analyzes the features of a person's voice delivered to a computer through a telephone, a terminal, or a microphone and finds the closest result from a pre-entered recognition list.
    [4] This speech recognition technology can be applied to a wide range of industries and applied as an interactive interface between people and machines, leading to high added value of products and contributing to the improvement of public welfare.
    [5] Voice recognition technology applied to a mobile terminal such as a mobile communication terminal or a portable information terminal should be able to optimally recognize the user's voice while having low power consumption due to the characteristics of the mobile terminal.
    [6] To this end, the mobile terminal adopts a speaker adaptation method that adapts to the user's voice, a small vocabulary that recognizes hundreds of words from tens of words, and an isolated word method that recognizes a specific word within a predetermined range to recognize the user's voice with a small amount of computation. Power consumption is reduced to a minimum.
    [7] A brief description of the conventional speech recognition technology will be performed. The speech recognition collects voice data of a user, extracts feature parameters, stores them in a memory of the terminal, and compares them with the parameters of the database-based existing speech to determine which speech.
    [8] Hidden markov models, dynamic time warping (DTW) algorithms, neural networks, and the like are used as speech recognition engines for comparing the similarity between voice data input by a user and parameters of existing speech.
    [9] As consumers' demand for multimedia use in mobile devices is increasing, the most important technical issue is power consumption.
    [10] The reason why the speech recognition rate is getting higher and still does not apply to the portable terminal is the huge amount of computation and the power consumption that is proportionally consumed. Accordingly, a lot of efforts are being made in terms of system to reduce power consumption.
    [11] Speech recognition used in the portable terminal records the user's voice to extract feature parameters. Here, when the user's voice is recorded through the microphone, the start point and end point of the voice are ambiguous.
    [12] That is, in order to improve the speech recognition rate, it is necessary to extract only the speech spoken by the user and calculate a feature parameter to recognize the speech. Other background noise, or if the beginning and end of the user's voice are cut off and the parameters are not properly extracted, the speech recognition rate is significantly reduced.
    [13] For example, a voice recognition technology applicable to a mobile terminal is to say a name that the user wants to find in the phonebook search menu, and the terminal performs voice recognition to find a desired name.
    [14] However, in the prior art as described above, when a user performs a function corresponding to speech recognition in a menu to use speech recognition, the mobile terminal does not know at what time the user speaks, and thus the sound continues through the microphone. Accept. In addition, even when the user finishes talking, the voice is received to a certain point to perform voice recognition. Thus, performing such an operation can be a significant fatal product defect in a mobile communication terminal where power consumption is important.
    [15] In addition, when there is a lot of noise or other voice sources around the terminal, even when the user finishes speaking, the user continuously receives the sound and performs the voice recognition in the recognition module. Therefore, there is a problem that consumes power unnecessarily.
    [16] Accordingly, the present invention has been made in view of the above problems, and by assigning an arbitrary key as a voice recognition key on the terminal keypad, the mobile communication terminal receives and recognizes the voice only while the user presses the voice recognition key. An object of the present invention is to provide a voice recognition method of a mobile communication terminal.
    [1] 1 is a flowchart illustrating a voice recognition method of a mobile communication terminal according to the present invention;
    [17] The present invention for achieving the above object, the step of assigning any key of the keypad as a key for speech recognition; Determining whether the user presses the voice recognition key and storing the signal input through the microphone and the analog / digital converter in a memory; Determining whether to stop the storing of the digitally converted signal in a memory by determining whether the voice recognition key falls; And extracting a feature parameter from the data stored in the memory to recognize the user's voice through a speech recognition engine.
    [18] Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.
    [19] 1 is a flowchart illustrating a method of recognizing a voice of a mobile communication terminal according to the present invention. As shown in FIG. 1, when a user presses a key for recognizing a voice, a signal input through a microphone and an analog / digital converter is determined. Storing them in a memory (S11, S12); Determining whether the voice recognition key falls and determining whether to stop storing the digitally converted signal in a memory (S13, S14); Extracting feature parameters from the data stored in the memory to recognize the user's voice through the speech recognition engine (S15, S16).
    [20] The present invention is configured to assign a voice recognition key to a keypad so as to receive and recognize the user's voice only when the voice recognition key is pressed.
    [21] When the user presses a key for voice recognition to use voice recognition, an interrupt occurs and the mobile communication terminal recognizes that the user wants to recognize the voice (S11).
    [22] The mobile communication terminal outputs a guide message to the user, that is, a message that the user's voice is recorded for voice recognition from now on. This process allows the user to speak more carefully and knows that the mobile terminal is trying to recognize the voice.
    [23] The mobile communication terminal converts the voice received through the microphone into digital data via an analog / digital converter and stores it in memory (S12).
    [24] When the user pronounces the voice to be recognized and releases the key for voice recognition, an interrupt is generated again, and the mobile communication terminal stops storing digital data coming through the analog / digital converter through the microphone (S13 and S14).
    [25] Thereafter, the mobile communication terminal extracts the feature parameter from the voice data recorded in the memory and compares the feature parameter with the parameter stored in the existing database (S15).
    [26] The mobile communication terminal that performs the voice recognition recognizes the user's voice and performs an operation corresponding to the voice.
    [27] As described in detail above, the present invention has an effect of increasing the battery life and enabling the speech recognition function to be universally used in the portable terminal by more efficiently using the speech recognition in the multimedia portable terminal, the power consumption is a major concern.
    权利要求:
    Claims (1)
    [1" claim-type="Currently amended] Assigning any key of the keypad as a key for speech recognition; Determining whether the user presses a key for speech recognition and storing the signal input through the microphone and the analog / digital converter in a memory; Determining whether to stop the storing of the digitally converted signal in a memory by determining whether the voice recognition key falls; Extracting feature parameters from the data stored in the memory and recognizing a user's voice through a voice recognition engine.
    类似技术:
    公开号 | 公开日 | 专利标题
    US10522152B2|2019-12-31|Diarization using linguistic labeling
    US9601114B2|2017-03-21|Method for embedding voice mail in a spoken utterance using a natural language processing computer system
    US10187503B2|2019-01-22|Enabling voice control of telephone device
    US10446140B2|2019-10-15|Method and apparatus for identifying acoustic background environments based on time and speed to enhance automatic speech recognition
    US10553216B2|2020-02-04|System and method for an integrated, multi-modal, multi-device natural language voice services environment
    US9183843B2|2015-11-10|Configurable speech recognition system using multiple recognizers
    US9892728B2|2018-02-13|System and method for mobile automatic speech recognition
    US9117449B2|2015-08-25|Embedded system for construction of small footprint speech recognition with user-definable constraints
    US8700397B2|2014-04-15|Speech recognition of character sequences
    US9666188B2|2017-05-30|System and method of performing automatic speech recognition using local private data
    US8204749B2|2012-06-19|System and method for building emotional machines
    US10643614B2|2020-05-05|Promoting voice actions to hotwords
    US9336773B2|2016-05-10|System and method for standardized speech recognition infrastructure
    Deng et al.2004|Challenges in adopting speech recognition
    KR101798828B1|2017-11-16|System and method for hybrid processing in a natural language voice services environment
    US8898065B2|2014-11-25|Configurable speech recognition system using multiple recognizers
    US8719020B1|2014-05-06|Generation of voice profiles
    JP2015018265A|2015-01-29|Speech recognition repair using contextual information
    US6463413B1|2002-10-08|Speech recognition training for small hardware devices
    US8639508B2|2014-01-28|User-specific confidence thresholds for speech recognition
    US8244540B2|2012-08-14|System and method for providing a textual representation of an audio message to a mobile device
    JP3363630B2|2003-01-08|Voice recognition method
    US7103542B2|2006-09-05|Automatically improving a voice recognition system
    US7603279B2|2009-10-13|Grammar update system and method for speech recognition
    JP6113302B2|2017-04-12|Audio data transmission method and apparatus
    同族专利:
    公开号 | 公开日
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    法律状态:
    2002-11-14|Application filed by 엘지전자 주식회사
    2002-11-14|Priority to KR1020020070894A
    2004-05-22|Publication of KR20040042942A
    优先权:
    申请号 | 申请日 | 专利标题
    KR1020020070894A|KR20040042942A|2002-11-14|2002-11-14|Voice recognition method for mobile communication terminal|
    [返回顶部]