New! View global litigation for patent families

JP2008224960A - Voice recognition device - Google Patents

Voice recognition device

Info

Publication number
JP2008224960A
JP2008224960A JP2007061947A JP2007061947A JP2008224960A JP 2008224960 A JP2008224960 A JP 2008224960A JP 2007061947 A JP2007061947 A JP 2007061947A JP 2007061947 A JP2007061947 A JP 2007061947A JP 2008224960 A JP2008224960 A JP 2008224960A
Authority
JP
Grant status
Application
Patent type
Prior art keywords
voice
unit
removal
recognition
characteristic
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP2007061947A
Other languages
Japanese (ja)
Inventor
Takeshi Ejiri
Machiko Ikeda
Shigeru Takahashi
Futoshi Tazaki
Toshio Tsukamoto
敏男 塚本
剛士 江尻
真知子 池田
太 田崎
高橋  滋
Original Assignee
Nippon Seiki Co Ltd
日本精機株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date

Links

Images

Abstract

PROBLEM TO BE SOLVED: To provide a voice recognition device for easily removing noise at low cost.
SOLUTION: A noise removing unit 110 removes noise included in a voice input from a microphone 102. A storage unit 120 stores removal characteristic parameters determining noise removal characteristics of the noise removing unit 110. A removal characteristic parameter determining unit 130 selects removal characteristic parameters out of the removal characteristic parameters stored in the storage unit 120 based upon surroundings when voice recognition is started and the voice quality of the speaker. The removal characteristic parameter determining unit 130 updates the removal characteristic parameters determined at the start of the voice recognition when a voice recognition unit 104 cannot obtain a voice recognition result.
COPYRIGHT: (C)2008,JPO&INPIT

Description

本発明は、マイクロフォンから入力した音声信号に含まれるノイズを除去し、音声を認識する音声認識装置に関するものである。 The present invention removes the noise contained in the speech signal input from microphone relates recognizing speech recognizer voice.

近年、カーナビゲーション装置や空調装置などの車載用機器を音声にて操作するための音声認識装置が種々提案されている。 In recent years, voice recognition device for operating a vehicle equipment such as car navigation devices and air-conditioning system in the voice have been proposed. 斯かる音声認識装置は、発話者の音声を取り込むためのマイクロフォンを備え、このマイクロフォンで取り込んだ音声信号から発話者の発話語句を認識するものである。 Such speech recognition device includes a microphone for capturing voice of a speaker, it is to recognize the speech phrase speaker from the voice signal captured by the microphone.
車両内においては、車両のエンジン音、空調装置の稼動音、走行によって発生する風音やロードノイズなど様々なノイズが多く存在するため、マイクロフォンを介して得られた発話者の音声信号にはノイズが重畳されてしまい、ノイズが重畳された音声に対して音声認識を行っても、十分な精度の認識が行えない場合が多い。 In the vehicle, engine sound of a vehicle, operation noise of the air conditioning apparatus, since there are many different noise such as wind noise and road noise generated by the running, the speaker of the audio signal obtained through the microphone noise There will be superimposed, even if the speech recognition on speech noise is superimposed, often not be recognized with sufficient accuracy. 従って、車両内のような騒音環境下において音声認識の精度を高めるためには、マイクロフォンで取り込んだ音声信号から車両内のノイズを除去する必要がある。 Therefore, in order to increase the accuracy of speech recognition in noise environments, such as in the vehicle, it is necessary to remove the noise in the vehicle from the audio signal captured by the microphone.
そこで、複数のマイクロフォンを異なる位置に配置して、夫々のマイクロフォンで収集した音声信号に基づき発話者の音声と雑音とを分離する方法や、発話音声の直前のノイズを分析して、発話中の音声に重畳されたノイズを差し引くことでノイズを除去する方法が知られている。 Therefore, by arranging a plurality of microphones at different positions, and a method for separating a speaker of the speech and noise based on the sound signals collected by the microphone of the respective analyzes the immediately preceding noise speech, in the speech method of removing noise by subtracting the superimposed noise speech is known. また、車両内のノイズは低周波成分が集中することが知られており、この性質に基づいて高域通過フィルタ、または帯域通過フィルタを用いたノイズ除去装置が提案されている(特許文献1〜4参照)。 The noise in the vehicle is known to the low-frequency component is concentrated, high-pass filter based on the nature or noise removal apparatus using the band-pass filter, has been proposed (Patent Documents 1 reference 4).
特許第3786038号公報 Patent No. 3786038 Publication 特許第3176474号公報 Patent No. 3176474 Publication 特開2001−296887号公報 JP 2001-296887 JP 特開2002−221986号公報 JP 2002-221986 JP

特許文献1,2に開示された音声認識装置は、ノイズ除去性能は高いが、複数のマイクロフォンが必要であり、処理が複雑であるが故に製造コストが高くなってしまう。 Speech recognition apparatus disclosed in Patent Documents 1 and 2, the noise removal performance is high but requires more microphones, processing becomes high is complex because the production cost. 特許文献3に開示された音声認識装置では、車両内の騒音の周波数特性が車種や走行状態によって変化することに着目して、発話音声直前の背景雑音を分析してノイズ除去のためのフィルタ特性を決定しており、車種や走行状態に関わらずノイズ除去のフィルタ特性を固定にした場合に比して、認識性能は向上するが、例えば発話者が低い声の男性の場合、発話者の音声信号がノイズに埋もれてしまい、ノイズが十分に除去されずに認識率が向上しない虞がある。 In the speech recognition apparatus disclosed in Patent Document 3, by focusing on the frequency characteristic of the noise in the vehicle changes depending on the model and the running state, the filter characteristics for analyzing the background noise immediately before speech denoising We are determined and compared with the case where the filter characteristics of the noise removal regardless models and the running state is fixed, the recognition performance is improved, for example, if the speaker is a low voice of man, speaker speech buried signal to noise, noise is likely that recognition rate is not improved without being sufficiently removed. 換言すると、ノイズ除去のフィルタ特性を決定する際に発話者の声質を考慮していないため、ノイズが十分に除去できない場合がある。 In other words, it does not take into account the voice of a speaker in determining the filter characteristic of the noise reduction, in some cases noise can not be sufficiently removed.

特許文献4に開示された音声認識装置では、音声認識に先んじて、ノイズレベルが低い環境下で認識対象となる幾つかの語句を発話者に発声してもらった音声信号と、所定の走行状態での車両内ノイズとを予め用意しておき、その音声信号と車両内ノイズとを合成した信号に対して最も良くノイズが除去できるフィルタ特性を見つけ出すことで、最適なノイズ除去のフィルタ特性を決定する方法を提案している。 The speech recognition apparatus disclosed in Patent Document 4, in advance of speech recognition, the speech signal several words the noise level is recognized at low environment were asked to utterance speaker, a predetermined running state prepared in advance and the vehicle in noise at, by finding the filter characteristics can be best noise removal and the audio signal and the in-vehicle noise to synthesized signal, determines the filter characteristics of the optimum noise elimination It has proposed a method of. この方法は、車種や走行状態だけでなく、発話者の声質も加味してフィルタ特性を決定できるため、文献3に開示された音声認識装置に比べて性能の高いノイズ除去装置が提供できる。 This method not only model and the running state, it is possible to determine the filter characteristics taking into account also the voice quality of a speaker, the performance of high noise removing device can be provided in comparison with the voice recognition device disclosed in the literature 3. しかし、事前に発話者に発声してもらい、走行時のノイズを録音する必要があるなど、ユーザに負担を強いる方法でもある。 However, prior asked to speaking to the speaker on, such as it is necessary to record the noise at the time of traveling, there is also a method imposes a burden on the user.
本発明は、安価で簡易な形態でノイズ除去が可能である音声認識装置を提供するものである。 The present invention is to provide a speech recognition apparatus is capable of removing noise by the simple form at low cost.

本発明は、請求項1に記載したように、マイクロフォン102から入力された音声に含まれるノイズを除去するノイズ除去部110と、前記ノイズ除去部110のノイズ除去特性を決定する除去特性パラメータを格納する記憶部120と、前記記憶部120に格納された前記除去特性パラメータの中から、音声認識を開始する時点での周辺状態と発話者の声質とに基づいて前記除去特性パラメータを選択する除去特性パラメータ決定部130と、を備え、前記ノイズ除去部110によってノイズが除去された音声信号を音声認識部104に入力する音声認識装置であって、前記除去特性パラメータ決定部130は、前記音声認識部104にて音声認識結果が得られなければ音声認識開始時に決定された前記除去特性パラメータを更新するもの The present invention as described in claim 1, storing a noise removing unit 110 for removing noise included in the voice input from the microphone 102, the removal characteristics parameters determining the noise rejection of the noise removing unit 110 removal characteristics and storage unit 120, to select from among the removal characteristic parameter stored in the storage unit 120, the removal characteristics parameters based on the voice quality of ambient conditions and the speaker at the time of starting the voice recognition to a parameter determination unit 130 includes a, an audio signal from which the noise is removed by the noise removing unit 110 a speech recognition apparatus for inputting the voice recognition unit 104, the removal characteristics parameter determination unit 130, the voice recognition unit 104 which updates the removal characteristic parameter speech recognition result has been determined at the speech recognition start to be obtained at ある。 A.

また、本発明は、請求項2に記載したように、前記周辺状態は車速データを含むものである。 The present invention, as set forth in claim 2, wherein said peripheral condition is intended to include vehicle speed data.

ノイズ除去部のノイズ除去特性を決定する除去特性パラメータは、音声認識開始時の周辺状態と発話者の声質に基づいて決定され、音声認識部に入力した音声信号が音声認識部に用意されたすべての語句に適合しなかった場合には、ノイズ除去性能を高めるように前記除去特性パラメータを更新するので、高性能なノイズ除去が可能になる。 Removal characteristic parameters for determining the noise rejection of the noise removing unit is determined based on the voice quality of ambient conditions and the speaker at the speech recognition start, all the audio signal is provided to the speech recognition unit which is input to the speech recognition unit of if not compatible with the phrase, so updates the removal characteristic parameter so as to enhance the noise rejection, allowing high-performance noise reduction. 且つ、前記除去特性パラメータを決定するにあたって、事前に発話者に発声してもらったり、走行時のノイズを録音したりする必要がない。 And, in determining the removal characteristic parameter in advance or asked to say a speaker to, there is no need or record noise during running.

以下、本発明の実施形態を添付図面に基づいて説明する。 It will be described below with reference to embodiments of the present invention in the accompanying drawings. 音声認識装置100は、ノイズ除去装置101と、発話者の音声をアナログの音声信号に変換するマイクロフォン102と、アナログ音声信号をデジタル信号に変換するA/D変換器103と、デジタル音声信号を入力して音声認識を行う音声認識部104と、車両状態の一要素である車速を電気信号として出力する車速パルス生成機105と、発話者の声の高さを設定するための声質選択スイッチ106とを備えている。 Input speech recognition apparatus 100 includes a noise removal device 101, a microphone 102 for converting a voice of a speaker into an analog audio signal, an A / D converter 103 for converting an analog audio signal into a digital signal, a digital audio signal a speech recognition unit 104 which performs speech recognition by a vehicle speed pulse generator 105 which outputs a vehicle speed which is an element of the vehicle state as an electric signal, a voice select switch 106 for setting the height of the voice of the speaker It is equipped with a.

ノイズ除去装置101は、ノイズ除去部110と、記憶部120と、除去特性パラメータ決定部130とで構成される。 The noise removal apparatus 101 includes a noise removing unit 110, a storage unit 120, and a removal characteristic parameter determining unit 130. ノイズ除去部110は、マイクロフォン102で捉えた車両内周辺ノイズを含む音声信号からノイズが集中する周波数帯を取り除くバンドパスフィルタ(BPF)111と、BPF111でノイズを除去した音声信号の音圧を調整する増幅器112とで構成される。 Noise removing unit 110, the adjustment band pass filter (BPF) 111 which removes the frequency band to concentrate noise from the audio signal including the in-vehicle ambient noise captured by the microphone 102, the sound pressure of the audio signal to remove noise BPF111 constituted by the amplifier 112 to. 記憶部120は、除去特性パラメータを多数格納するものであり、例えばフラッシュROMのように書換え可能で不揮発性の記憶媒体からなる。 Storage unit 120 is for storing a large number of rejection parameters, for example, a rewritable nonvolatile storage medium such as a flash ROM. 除去特性パラメータ決定部130は、除去特性パラメータの決定、更新及びノイズ除去部の制御を行う中央演算装置(CPU)131と、前記処理のプログラムが格納されたROM132と、演算結果を保持するRAM133とで構成される。 Removal characteristic parameter determining unit 130 determines the removal characteristic parameters, a central processing unit (CPU) 131 for controlling the updating and the noise removing unit, a ROM132 that the processing of the program is stored, a RAM133 that holds the operation result in constructed.

次に、図2に基づいて、除去特性パラメータを更新する手順を説明する。 Next, based on FIG. 2, illustrating a procedure for updating the rejection parameters. まず、スイッチ押下などによって音声認識の開始を検知するところから処理が始まる。 First, processing where it detects the start of the voice recognition by a switch pressing begins. 音声認識の開始を検知したら、CPU131は音声認識開始以前までに車速パルス生成機105が送出する車速パルスをもとに音声認識開始時点での車速データを算出し、図3に示す表に基づいて、車速データから車両状態区分を決定する(S201)。 Upon detecting the start of speech recognition, CPU 131 calculates the vehicle speed data at the speech recognition start point based on the vehicle speed pulse to be sent is the vehicle speed pulse generator 105 until the voice recognition start earlier, based on the table shown in FIG. 3 to determine the vehicle condition classification from the vehicle speed data (S201). また、CPU131は、ユーザによって設定された声質選択スイッチ106の状態に基づいて、声質種別を決定する(S202)。 Further, CPU 131, based on the state of the voice selection switch 106 which is set by the user, determines the voice type (S202). 発話者の声の高さは高/中/低の3段階で区分されており、声質選択スイッチ106が3段階のうちのいずれか1つを選択でき、CPU131は性質選択スイッチ106の状態に応じて、音質種別を高/中/低のいずれかに決定する。 Voice pitch of a speaker is classified in a high / medium / low three levels, you can select any one of the voice selection switch 106 is three steps, CPU 131 is depending on the state of nature selection switch 106 Te, to determine the sound quality type in the high / medium / low one of the.

CPU131は、ステップS201及びステップS202で決定された車両状態区分と声質種別とに対応する除去特性パラメータを記憶部120から読み込む(S203)。 CPU131 reads removal characteristics parameters corresponding to the vehicle state classification and voice type and determined in step S201 and step S202 from the storage unit 120 (S203). なお、除去特性パラメータは、BPF111の低域側カットオフ周波数fcl、高域側カットオフ周波数fch及び増幅器112の利得Aの3つの要素からなり、図4に示すように、車両状態区分・声質種別と対応付けられて記憶部120に格納されている。 Incidentally, rejection parameters, low frequency side cutoff frequency fcl of the BPF 111, consists of three elements of the high frequency side cutoff frequency fch and gain A of the amplifier 112, as shown in FIG. 4, the vehicle state classification - voice classification stored in the storage unit 120 in association with.

ステップS203で決定された除去特性パラメータに基づき、CPU131はBPF111、増幅器112の動作特性を変更する。 Based on removal characteristic parameters determined in step S203, CPU 131 may BPF 111, to change the operating characteristics of the amplifier 112. なお、例えばBPF111がバターワース型IIRフィルタで構築される場合、BPF111を一意に決定するためにはタップ数や各タップにおける重み付け係数を決定する必要があるが、これらのパラメータに関し前記除去特性パラメータのみから算出できないものについては、事前に制約を設けておく。 Incidentally, for example, if the BPF111 is constructed by Butterworth IIR filter, but in order to uniquely determine the BPF111 is necessary to determine the weighting factors in the number of taps and the tap, only the removing characteristic parameter relates to these parameters for those that can not be calculated in advance in advance to provide a constraint.

ノイズ除去部110のノイズ除去特性が設定された後、マイクロフォン102とA/D変換器103とを通じて発話音声を取り込み、ノイズ除去部110を介して、取り込んだ発話音声を音声認識部104へ入力すると同時に、CPU131は発話音声のデジタルデータをRAM133に格納する(S205)。 After the noise removal characteristic of the noise removing unit 110 is set, it captures speech through a microphone 102 and an A / D converter 103, via the noise removing unit 110, entering a speech taken to the speech recognition unit 104 At the same time, CPU 131 stores the digital data of the speech to the RAM 133 (S205).
次に、CPU131は音声認識部104から認識結果を受け取り、認識結果が得られれば(即ち、音声認識部104に登録されている語句中のいずれかの語句が得られれば)、ステップ215へ進む(S207)。 Next, CPU 131 receives the recognition result from the voice recognition unit 104, as long obtained recognition results (i.e., as long obtained either language in the word registered in the speech recognition unit 104), the process proceeds to step 215 (S207).

ステップS207において、音声認識部104に登録されている語句中のいずれの語句も音声認識部104が出力しない場合(即ち、認識結果が得られない場合)、まず始めにBPF111の低域側カットオフ周波数fclを低域側カットオフ周波数の上限Fclと比較する(S208)。 In step S207, if none of the words in the word registered in the speech recognition unit 104 does not output the speech recognition unit 104 (i.e., if the recognition result is not obtained), firstly to the low frequency side cutoff BPF111 frequency fcl compared with the upper limit fcl lower frequency cut-off frequency (S208). 低域側カットオフ周波数fclが低域側カットオフ周波数の上限Fclより小さい場合、低域側カットオフ周波数fclを段階的に引き上げる(S209)。 When the low frequency side cutoff frequency fcl upper limit Fcl less than the low frequency side cutoff frequency stepwise raising the low frequency side cutoff frequency fcl (S209). 一方、低域側カットオフ周波数fclが低域側カットオフ周波数の上限Fcl以上である場合は、次に高域側カットオフ周波数fchと高域側カットオフ周波数の下限Fchとを比較し(S210)、高域側カットオフ周波数fchが高域側カットオフ周波数の下限Fchより大きい場合は高域側カットオフ周波数fchを段階的に引き下げ(S211)、高域側カットオフ周波数fchが高域側カットオフ周波数の下限Fch以下である場合はステップS215へ進む。 On the other hand, when the low frequency side cutoff frequency fcl is the upper limit Fcl more lower frequency side cutoff frequency, then compares the lower limit Fch the high frequency side cutoff frequency fch and the high frequency side cutoff frequency (S210 ), the high frequency side cutoff frequency fch stepwise lowered the high frequency side cutoff frequency fch if greater than the lower limit fch high frequency side cutoff frequency (S211), the high frequency side cutoff frequency fch high frequency side If it is less than the lower limit Fch cutoff frequency, the process proceeds to step S215.

低域側カットオフ周波数fcl、または高域側カットオフ周波数fchのいずれか一方が変更されたならば、増幅器112の利得Aをfcl、fchに基づき変更する(S212)。 If the low frequency side cutoff frequency fcl or either of the high frequency side cutoff frequency fch, is changed, the gain A of the amplifier 112 fcl, is changed based on the fch (S212). その後、再度ノイズ除去部のノイズ除去特性を設定し直し(S213)、CPU131はステップS205でRAM133に格納した音声信号を、ノイズ除去部110を介して音声認識部104に入力する(S214)。 Then, reset the noise rejection of the re-noise removing unit (S213), CPU 131 is an audio signal stored, in step S205 RAM 133, is input to the speech recognition unit 104 through the noise removing unit 110 (S214).
ステップS206からステップS214までを、fcl、fch、Aが変更されなくなるまで繰り返し、fcl、fch、Aが変更されなくなったらステップS201及びステップS202で決定した車両状態区分、声質種別に対応する除去特性パラメータを記憶部120へ書き込む(S215)。 Steps S206 to step S214, fcl, fch, A repeated until no change, fcl, fch, vehicle state classification determined in step S201 and S202 When A is no longer changed, removed characteristic parameter corresponding to the voice quality classification the writing into the storage unit 120 (S215).

なお、本実施形態では車両状態を車速データで決定したが、本発明では車両状態が車速で一意に決定されるものではなく、例えばエンジン回転数やワイパーの駆動有無などの車両情報の組合せで決定するものであっても良い。 Incidentally, determining the vehicle state in the present embodiment has been determined by the vehicle speed data, and not the vehicle state is uniquely determined by the vehicle speed in the present invention, for example, a combination of vehicle information such as driving whether the engine rotational speed and the wiper it may be one that.
なお、本実施形態では発話者の声質種別をスイッチの状態で決定したが、本発明においては、例えば車両内ノイズが小さい場合の音声を分析して、発話音声の基本周波数などの音声特徴を捉え、その音声特徴をもとに声質を符号化して声質種別を決定するものであっても良い。 In the present embodiment has been determined voice type of a speaker in the state of the switch, in the present invention, for example by analyzing a voice in case the vehicle in the noise is small, it captures the speech features such as the fundamental frequency of the speech , it may be configured to determine a voice type encodes the voice based on the voice features.

また、本実施形態ではノイズ除去装置101と音声認識部104が別の装置であったが、ノイズ除去も音声認識もソフトウェアとして構成して一つの装置で処理するものであっても良い。 Further, in the present embodiment the noise removal apparatus 101 and the speech recognition unit 104 is a separate device, or may be treated with one device constitutes noise removing both speech recognition as software. また、本実施形態ではノイズ除去装置101がA/D変換器103の後段に設置してあったが、ノイズ除去部110をアナログ回路で構成してA/D変換器103の前段に設置するものであっても良い。 Moreover, those noise removal device 101 in the present embodiment but had been installed in the rear stage of the A / D converter 103, a noise removing unit 110 and an analog circuit is placed in front of the A / D converter 103 it may be.

本発明の実施形態を示すブロック図。 Block diagram illustrating an embodiment of the present invention. 同上実施形態を示すフロー図。 Flow diagram showing the same embodiment. 同上実施形態を示す車速データと車両状態区分の関係を表す図。 Diagram showing the relationship between vehicle speed data and vehicle status classification showing the same embodiment. 同上実施形態を示す除去特性パラメータを表す図。 Diagram illustrating removal characteristic parameter showing the same embodiment.

符号の説明 DESCRIPTION OF SYMBOLS

102 マイクロフォン 104 音声認識部 110 ノイズ除去部 120 記憶部 130 除去特性パラメータ決定部 102 microphone 104 voice recognition unit 110 a noise removing unit 120 storage unit 130 removing characteristic parameter determination unit

Claims (2)

  1. マイクロフォンから入力された音声に含まれるノイズを除去するノイズ除去部と、前記ノイズ除去部のノイズ除去特性を決定する除去特性パラメータを格納する記憶部と、前記記憶部に格納された前記除去特性パラメータの中から、音声認識を開始する時点での周辺状態と発話者の声質とに基づいて前記除去特性パラメータを選択する除去特性パラメータ決定部と、を備え、前記ノイズ除去部によってノイズが除去された音声信号を音声認識部に入力する音声認識装置であって、前記除去特性パラメータ決定部は、前記音声認識部にて音声認識結果が得られなければ音声認識開始時に決定された前記除去特性パラメータを更新することを特徴とする音声認識装置。 A noise removing unit for removing noise contained in the speech input from the microphone, a storage unit for storing the removed characteristic parameter for determining the noise rejection of the noise removing unit, the removal characteristics parameters stored in the storage unit among comprises a removal properties parameter determining section for selecting the removing characteristic parameters based on the voice quality of ambient conditions and the speaker at the time of starting the voice recognition, a noise is removed by the noise removing unit a speech recognition apparatus for inputting an audio signal to the voice recognition unit, said removal characteristic parameter determination unit, the removal characteristic parameter speech recognition result has been determined at the speech recognition start to be obtained by the voice recognition unit speech recognition apparatus characterized by update.
  2. 前記周辺状態は車速データを含むことを特徴とする請求項1に記載の音声認識装置。 Speech recognition apparatus according to claim 1 wherein the peripheral state, characterized in that it comprises a vehicle speed data.
JP2007061947A 2007-03-12 2007-03-12 Voice recognition device Pending JP2008224960A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2007061947A JP2008224960A (en) 2007-03-12 2007-03-12 Voice recognition device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2007061947A JP2008224960A (en) 2007-03-12 2007-03-12 Voice recognition device

Publications (1)

Publication Number Publication Date
JP2008224960A true true JP2008224960A (en) 2008-09-25

Family

ID=39843671

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2007061947A Pending JP2008224960A (en) 2007-03-12 2007-03-12 Voice recognition device

Country Status (1)

Country Link
JP (1) JP2008224960A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101561098B1 (en) 2014-09-11 2015-10-16 현대자동차주식회사 vehicle and control method of the same

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000029500A (en) * 1998-07-08 2000-01-28 Denso Corp Noise suppressing apparatus and voice recognizing system using the apparatus
JP2002287789A (en) * 2001-03-28 2002-10-04 Alpine Electronics Inc Noise reduction system
JP2002314637A (en) * 2001-04-09 2002-10-25 Denso Corp Device for reducing noise
JP2004212641A (en) * 2002-12-27 2004-07-29 Toshiba Corp Voice input system and terminal device equipped with voice input system
JP2005164988A (en) * 2003-12-03 2005-06-23 Xanavi Informatics Corp Frequency switching device and information processing apparatus
JP2006163231A (en) * 2004-12-10 2006-06-22 Internatl Business Mach Corp <Ibm> Device, program, and method for noise elimination

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000029500A (en) * 1998-07-08 2000-01-28 Denso Corp Noise suppressing apparatus and voice recognizing system using the apparatus
JP2002287789A (en) * 2001-03-28 2002-10-04 Alpine Electronics Inc Noise reduction system
JP2002314637A (en) * 2001-04-09 2002-10-25 Denso Corp Device for reducing noise
JP2004212641A (en) * 2002-12-27 2004-07-29 Toshiba Corp Voice input system and terminal device equipped with voice input system
JP2005164988A (en) * 2003-12-03 2005-06-23 Xanavi Informatics Corp Frequency switching device and information processing apparatus
JP2006163231A (en) * 2004-12-10 2006-06-22 Internatl Business Mach Corp <Ibm> Device, program, and method for noise elimination

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101561098B1 (en) 2014-09-11 2015-10-16 현대자동차주식회사 vehicle and control method of the same

Similar Documents

Publication Publication Date Title
US6959276B2 (en) Including the category of environmental noise when processing speech signals
US20070078649A1 (en) Signature noise removal
US20100088093A1 (en) Voice Command Acquisition System and Method
US20070010997A1 (en) Sound processing apparatus and method
JP2000347688A (en) Noise suppressor
JP2008236448A (en) Sound signal processing device, hands-free calling device, sound signal processing method, and control program
US20050213776A1 (en) Automobile audio system and signal processing method therefor
JPH09232896A (en) Audio signal processing method
JP2002099297A (en) Microphone device
JP2006208482A (en) Device, method, and program for assisting activation of conference, and recording medium
JP2007318528A (en) Directional sound collector, directional sound collecting method, and computer program
WO2005013262A1 (en) Method for driving a dialog system
US5878392A (en) Speech recognition using recursive time-domain high-pass filtering of spectral feature vectors
US20060031067A1 (en) Sound input device
JPH06236196A (en) Method and device for voice recognition
JP2009058708A (en) Voice processing system, method and program
JP2005203981A (en) Device and method for processing acoustic signal
WO2012097150A1 (en) Automotive sound recognition system for enhanced situation awareness
US20120062391A1 (en) Vehicle External Warning Sound Generation System and Method
JP2005289354A (en) Engine sound processing device
US20040236571A1 (en) Subband method and apparatus for determining speech pauses adapting to background noise variation
WO2005106843A1 (en) Reproduction control of an audio signal based on musical genre classification
JP2000163098A (en) Voice recognition device
JP2002258882A (en) Voice recognition system and information recording medium
JP2002351488A (en) Noise canceller and on-vehicle system

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20100120

A131 Notification of reasons for refusal

Effective date: 20120516

Free format text: JAPANESE INTERMEDIATE CODE: A131

A02 Decision of refusal

Effective date: 20120927

Free format text: JAPANESE INTERMEDIATE CODE: A02