JPH0519796A - Excitation signal encoding and decoding method for voice - Google Patents
Excitation signal encoding and decoding method for voiceInfo
- Publication number
- JPH0519796A JPH0519796A JP3167124A JP16712491A JPH0519796A JP H0519796 A JPH0519796 A JP H0519796A JP 3167124 A JP3167124 A JP 3167124A JP 16712491 A JP16712491 A JP 16712491A JP H0519796 A JPH0519796 A JP H0519796A
- Authority
- JP
- Japan
- Prior art keywords
- code
- periodicity
- vector
- frame
- codebook
- 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.)
- Granted
Links
Abstract
Description
【０００１】[0001]
【産業上の利用分野】この発明は雑音符号帳を用い、符
号駆動線形予測符号化、ベクトル和駆動線形予測符号化
に適用され、音声の信号系列を少ない情報量でデジタル
符号化する高能率音声符号化方法、その復号化方法に関
する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is applied to code-driven linear predictive coding and vector sum-driven linear predictive coding using a noise codebook, and is a highly efficient voice for digitally encoding a voice signal sequence with a small amount of information. The present invention relates to an encoding method and a decoding method thereof.
【０００２】[0002]
【従来の技術】ディジタル移動無線通信方式で電波を効
率的に利用し、また音声蓄積サービスで記憶媒体を効率
的に利用するために、高能率音声符号化方法が用いられ
ている。現在、音声を高能率に符号化する方法として、
原音声をフレームと呼ばれる５〜５０ms程度の一定間隔
の区間に分割し、その１フレームの音声を周波数スペク
トルの包絡形状と、その包絡形状に対応する線形フィル
タを駆動するための駆動音源信号という２つの情報に分
離し、それぞれを符号化することが提案されている。そ
の場合、駆動音源信号を符号化する方法として、駆動音
源信号を音声の基本周波数（ピッチ周期）に対応すると
考えられる周期成分と、それ以外の成分（言い換えれば
非周期成分）とに分離して符号化する方法が知られてい
る。この駆動音源情報の符号化法として符号駆動線形予
測符号化（Code-Excited Linear Prediction Coding:CE
LP）およびベクトル和駆動線形予測符号化(Vector Sum
Excited Linear Prodiction Coding:VSELP）法がある。
それぞれの技術については、M.R.Schroeder and B.S.At
al : "Code-ExcitedLinear Prediction(CELP) :High-qu
ality Speech at Very Low Bit Rates", Proc.ICASSP'8
5,25.1.1,pp.937-940,1985 ，およびI.A.Gerson and M.
A.Jasiuk :"Vector Sum Excited Linear Prediction (V
SELP) Speech Coding at 8 kbps", Proc. ICASSP'90,S
9.3,pp.461-464,1990、に述べられている。2. Description of the Related Art A high-efficiency voice encoding method is used to efficiently use radio waves in a digital mobile radio communication system and to efficiently use a storage medium in a voice storage service. Currently, as a method for highly efficient speech coding,
The original voice is divided into sections called frames, which have a constant interval of about 5 to 50 ms, and the voice of one frame is called the envelope shape of the frequency spectrum and the driving sound source signal for driving the linear filter corresponding to the envelope shape. It has been proposed to separate the information into two pieces and encode each piece. In that case, as a method of encoding the driving sound source signal, the driving sound source signal is separated into a periodic component considered to correspond to the fundamental frequency (pitch period) of the sound and other components (in other words, non-periodic component). A method of encoding is known. Code-Excited Linear Prediction Coding (CE)
LP) and vector sum driven linear predictive coding (Vector Sum
Excited Linear Prodiction Coding (VSELP) method.
MR Schroeder and BSAt
al: "Code-ExcitedLinear Prediction (CELP): High-qu
ality Speech at Very Low Bit Rates ", Proc.ICASSP'8
5,25.1.1, pp.937-940,1985, and IAGerson and M.
A. Jasiuk: "Vector Sum Excited Linear Prediction (V
SELP) Speech Coding at 8 kbps ", Proc. ICASSP'90, S
9.3, pp.461-464, 1990.
【０００３】これらの符号化方法は、図３に示すよう
に、入力端子１１に入力された原音声について音声分析
部１２において、その周波数スペクトルの包絡形状を表
すパラメータが計算される。この分析には通常、線形予
測法が用いられる。その線形予測パラメータは線形予測
パラメータ符号化部１３で符号化され、その符号化出力
は分岐され、線形予測パラメータ復号化部１４で復号化
され、その復号化された線形予測パラメータが線形予測
合成フィルタ１５のフィルタ係数として設定される。In these encoding methods, as shown in FIG. 3, a parameter representing the envelope shape of the frequency spectrum of the original speech input to the input terminal 11 is calculated in the speech analysis section 12. Linear prediction methods are typically used for this analysis. The linear prediction parameter is coded by the linear prediction parameter coding unit 13, the coded output is branched, and decoded by the linear prediction parameter decoding unit 14, and the decoded linear prediction parameter is the linear prediction synthesis filter. It is set as a filter coefficient of 15.
【０００４】適応符号帳１６において直前の過去の駆動
音源ベクトルをある周期（ピッチ周期）に相当する長さ
で切り出し、その切り出したベクトルをフレームの長さ
になるまで繰り返し、音声の周期成分と対応する時系列
符号ベクトルの候補が出力される。また雑音符号帳１
７，１８から音声の非周期成分と対応する時系列符号ベ
クトルの候補が出力される。雑音符号帳１７，１８は図
４に示すように通常白色ガウス性雑音を基調とし、１フ
レーム分の長さの各種の符号ベクトルが入力音声とは独
立にあらかじめ記憶されている。In the adaptive codebook 16, the immediately preceding driving excitation vector is cut out at a length corresponding to a certain cycle (pitch cycle), and the cut out vector is repeated until the length of the frame is reached, and the vector corresponds to the cycle component of the voice. A candidate of the time series code vector to be output is output. Also, the noise codebook 1
7 and 18 output time-series code vector candidates corresponding to the aperiodic component of the voice. As shown in FIG. 4, the noise codebooks 17 and 18 are based on normal white Gaussian noise, and various code vectors having a length of one frame are stored in advance independently of the input speech.
【０００５】適応符号帳１６，雑音符号帳１７，１８か
らの各時系列ベクトルの候補は重みつき加算部１９にお
いて、それぞれ乗算部２１1 ，２１2 ，２１3 で重みｇ
1 ，ｇ2 ，ｇ3 が乗算され、これら乗算出力は加算部２
２で加算される。この加算出力は駆動音源ベクトルとし
て線形予測合成フィルタ１５へ供給され、合成フィルタ
１５から合成（再生）音声が出力される。この合成音声
の入力端子１１からの原音声に対する歪みが距離計算部
２３で計算され、その計算結果に応じて符号帳検索部２
４により、適応符号帳１６における切り出し長さをかえ
た候補が選択され、かつ雑音符号帳１７，１８から他の
符号ベクトルが選択され、さらに重みつき加算部１９の
重みｇ1 ，ｇ2 ，ｇ3 が変更され、距離計算部２３で計
算された歪みが最小になるようにされる。歪み最小とな
ったときの適応符号帳１６の切り出し長を示す周期符号
と、雑音符号帳１７，１８の各符号ベクトルを示す雑音
符号と、重みｇ1 ，ｇ2 ，ｇ3 を示す重み符号と、線形
予測パラメータ符号とが符号化出力として出力され、伝
送または蓄積される。The time series vector candidates from the adaptive codebook 16 and the noise codebooks 17 and 18 are weighted by the weighting adder 19 in the multipliers 21 1 , 21 2 and 21 3 , respectively.
1 , g 2 and g 3 are multiplied, and these multiplication outputs are added by the adder 2
It is added by 2. This addition output is supplied to the linear prediction synthesis filter 15 as a driving sound source vector, and synthesis (playback) voice is output from the synthesis filter 15. The distortion of the synthesized speech with respect to the original speech from the input terminal 11 is calculated by the distance calculation section 23, and the codebook search section 2 is calculated according to the calculation result.
4, a candidate whose cutout length is changed in the adaptive codebook 16 is selected, and another code vector is selected from the noise codebooks 17 and 18, and the weights g 1 , g 2 , and g of the weighted addition unit 19 are selected. 3 is changed so that the distortion calculated by the distance calculator 23 is minimized. A periodic code indicating the cutout length of the adaptive codebook 16 when the distortion becomes minimum, a noise code indicating each code vector of the noise codebooks 17 and 18, and a weight code indicating weights g 1 , g 2 and g 3. , Linear prediction parameter code and are output as encoded outputs and are transmitted or stored.
【０００６】復号化は図５に示すように入力された線形
予測パラメータ符号が線形予測パラメータ復号化部２６
で復号化され、その予測パラメータが線形予測合成フィ
ルタ２７にフィルタ係数として設定される。それまでに
得られた直前の過去の駆動音源ベクトルと、入力された
周期符号とを用いて適応符号帳２８からその周期で過去
の駆動音源ベクトルを切り出し、これをフレーム分繰り
返した時系列符号ベクトルが出力され、また入力された
雑音符号が示す符号ベクトルが雑音符号帳２９，３１か
らそれぞれ時系列ベクトルとして読み出される。これら
時系列ベクトルは重みつき加算部３２で入力された重み
符号に応じて、それぞれ重み付けがなされた後、加算さ
れ、その加算出力が駆動音源ベクトルとして合成フィル
タ２７へ供給され、合成フィルタ２７から再生音声が得
られる。For decoding, the linear prediction parameter code input as shown in FIG.
And the prediction parameter is set in the linear prediction synthesis filter 27 as a filter coefficient. A time-series code vector obtained by cutting out the past drive excitation vector from the adaptive codebook 28 in that cycle using the immediately preceding drive excitation vector obtained thus far and the input periodic code, and repeating this for a frame. Is output, and the code vector indicated by the input random code is read from the random code books 29 and 31 as a time series vector. These time-series vectors are weighted according to the weight code input by the weighted addition unit 32, and then added, and the added output is supplied to the synthesis filter 27 as a driving sound source vector and reproduced from the synthesis filter 27. Voice is obtained.
【０００７】雑音符号帳２９，３１は符号化に用いられ
た雑音符号帳１７，１８と同一のものとされる。雑音符
号帳は１個のみ、あるいはさらに多くのものが用いられ
ることもある。符号駆動線形予測符号化においては、雑
音符号帳には、候補となるべきすべての符号ベクトルが
直接記憶されてある。つまり、候補となるべき符号ベク
トルの数がＮならば、雑音符号帳に記憶されている符号
ベクトルの数もＮである。The random codebooks 29 and 31 are the same as the random codebooks 17 and 18 used for encoding. Only one random codebook or more may be used. In code-driven linear predictive coding, all code vectors to be candidates are directly stored in the random codebook. That is, if the number of code vectors to be candidates is N, the number of code vectors stored in the random codebook is also N.
【０００８】ベクトル和駆動線形予測符号化では、雑音
符号帳は図６に示すように、記憶されているすべての符
号ベクトル（基本ベクトルと呼ぶ）が同時に読み出さ
れ、乗算部３３1 〜３３M でそれぞれ雑音符号帳用復号
器３４により＋１または−１が乗算され、その乗算出力
が加算されて出力符号ベクトルとして出力される。従っ
て、各基本ベクトルに乗算する＋１，−１の組み合わせ
により、出力符号ベクトルの数は２M となり、歪みが最
小となるようにこの２M の出力符号ベクトルの１つが選
択される。[0008] In the vector sum excited linear predictive coding, the noise codebook, as shown in FIG. 6, (referred to as a basic vector) all code vectors stored are read out simultaneously, multiplying unit 33 1 ~ 33 M Then, each of them is multiplied by +1 or -1 by the noise codebook decoder 34, and the multiplication outputs are added and output as an output code vector. Therefore, the number of output code vectors becomes 2 M by the combination of +1, −1 by which each basic vector is multiplied, and one of the 2 M output code vectors is selected so that the distortion is minimized.
【０００９】ところが、これらの従来の方法では、駆動
音源信号の周期性が前フレームの成分のみに限定される
ため、周期性の表現力が弱く、再生音声がざらざらして
滑らかさに欠けるという欠点を有していた。このような
点から、音声の周期性の表現力を強化するため、従来周
期性をもたなかった雑音符号帳から出力される符号ベク
トルの一部または全部、あるいは出力される符号ベクト
ルの成分の一部、もしくは複数の雑音符号帳の一部に適
応符号帳の出力時系列符号ベクトルの周期性と同一の周
期性をもたせることを提案した。However, in these conventional methods, since the periodicity of the driving sound source signal is limited only to the component of the previous frame, the expressiveness of the periodicity is weak and the reproduced voice is rough and lacks smoothness. Had. From this point of view, in order to enhance the expressiveness of the periodicity of the speech, a part or all of the code vector output from the noise codebook, which has no conventional periodicity, or the component of the output code vector, is used. It is proposed that some or some noise codebooks have the same periodicity as the output time-series codevector of the adaptive codebook.
【００１０】つまり図７に示すように、雑音符号帳１７
から１つの符号ベクトルを、基本周期検索（適応符号１
６の検索）で得られた基本周期Ｌの長さ分３６を切り出
す。ａに示すように、その切り出し部分３６をフレーム
長に達するまで何度も繰り返し配列して、周期性符号ベ
クトルを作成して出力符号ベクトルとする。それを雑音
符号帳１７中のすべての符号ベクトルについて行い、そ
の中で、合成フィルタに通した再生音声と原音声間の距
離が最小になるものを、最適符号ベクトルとする。その
後の各駆動音源成分の重みの決定は従来の技術と同様に
行う。復号側でもそれまでに得られたピッチ周期で雑音
符号帳の符号ベクトルを周期化する。That is, as shown in FIG. 7, the random codebook 17
From the basic period search (adaptive code 1
The length 36 of the basic period L obtained in (6 search) is cut out. As shown in a, the cut-out portions 36 are repeatedly arranged until the frame length is reached, and a periodic code vector is created to be an output code vector. This is performed for all code vectors in the noise codebook 17, and the code vector in which the distance between the reproduced voice and the original voice passed through the synthesis filter is minimized is set as the optimum code vector. Subsequent determination of the weight of each driving sound source component is performed in the same manner as the conventional technique. Also on the decoding side, the code vector of the random codebook is made periodic with the pitch period obtained so far.
【００１１】[0011]
【発明が解決しようとする課題】このように、雑音符号
帳の符号ベクトルもピッチ周期で繰り返す周期化処理を
行うことで、量子化歪が小さくなり、品質が改善される
が、必ずしもそうでないことがわかった。つまり周期性
の低いフレームに対しては雑音符号ベクトルを周期化す
ることは逆効果を生ずる場合があることが判明した。As described above, by performing the periodic process of repeating the code vector of the random codebook at the pitch period, the quantization distortion is reduced and the quality is improved, but this is not always the case. I understood. In other words, it was found that periodicization of the random code vector may have an adverse effect on a frame with low periodicity.
【００１２】[0012]
【課題を解決するための手段】この発明によれば、雑音
符号帳の符号ベクトルをピッチ周期と対応して適応させ
て周期化処理を行うが、現在または過去の音声の周期性
の程度を推定し、その推定した周期性の程度に応じてフ
レームごとに雑音符号帳の符号ベクトルの周期化の程
度、つまり周期化の数、または１つの符号ベクトルにつ
いての周期化の漏れを適応的に変更する。According to the present invention, the periodicization process is performed by adapting the code vector of the noise codebook in correspondence with the pitch period, but the degree of periodicity of the present or past speech is estimated. Then, according to the estimated degree of periodicity, the degree of periodicization of the code vector of the noise codebook, that is, the number of periodicizations, or the leakage of periodicity for one code vector is adaptively changed for each frame. .
【００１３】復号側においても符号化側と同様に処理す
る。The decoding side performs the same processing as the encoding side.
【００１４】[0014]
【実施例】図１にこの発明の実施例の要部を示す。適応
符号帳１６に対する基本周期の検索は従来と同様であ
る。この発明においては、周期化処理個数決定部３７で
現フレームの周期性の程度を推定する。この推定は例え
ばパワ、ピッチ利得、スペクトル包絡の形状によって周
期性の程度を推定し、その程度に応じて雑音符号帳１７
中の周期化処理を行う符号ベクトルの個数を決める。図
２Ａに示すように、推定した周期性が高い場合はピッチ
周期Ｌで周期化する符号ベクトルの数を多くし、推定周
期性が低い場合はピッチ周期Ｌで周期化処理する符号ベ
クトルの数を少なくする。雑音符号帳１７の最適符号ベ
クトルを決定する前にピッチ利得を独立に量子化する場
合には、周期性の推定としてピッチ利得を用い、ピッチ
利得にほぼ比例させて周期化する符号ベクトルの個数を
決める。またピッチ利得を雑音符号ベクトルの符号決定
後に、その雑音符号ベクトルの利得と同時に決定する場
合には、スペクトル包絡の傾きとパワを推定周期性とし
て用いる。ピッチの周期性の強弱はパワおよびスペクト
ル包絡の傾き（１次の予測係数）と強い相関があるた
め、ピッチの周期性の強さを推定できる。DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows the essential parts of an embodiment of the present invention. The search for the basic period in the adaptive codebook 16 is the same as the conventional one. In the present invention, the periodicity processing number determination unit 37 estimates the degree of periodicity of the current frame. In this estimation, for example, the degree of periodicity is estimated by the power, the pitch gain, and the shape of the spectrum envelope, and the noise codebook 17 is estimated according to the degree.
The number of code vectors to be subjected to the periodical processing is determined. As shown in FIG. 2A, when the estimated periodicity is high, the number of code vectors to be periodicized with the pitch period L is increased, and when the estimated periodicity is low, the number of code vectors to be periodicized with the pitch period L is increased. Reduce. When the pitch gain is independently quantized before the optimum code vector of the noise codebook 17 is determined, the pitch gain is used as the estimation of the periodicity, and the number of code vectors to be periodicized in proportion to the pitch gain is set. Decide When the pitch gain is determined at the same time as the gain of the noise code vector after the code of the noise code vector is determined, the slope of the spectrum envelope and the power are used as the estimated periodicity. Since the strength of the pitch periodicity has a strong correlation with the power and the slope of the spectrum envelope (first-order prediction coefficient), the strength of the pitch periodicity can be estimated.
【００１５】現フレームについて周期性を推定するほか
に、過去に既に復号化された音声のもつピッチの周期性
を利用してもよい。つまり復号化された音声は符号器、
復号器で共通に得られ、またピッチの周期性は連続する
フレームでは急変しない性質があるため、過去のフレー
ムの周期性を推定してもよい。この過去のフレームの周
期性の推定は、例えば自己相関関数の大きさから推定す
る。これまでの説明では周期性の推定を従来の符号化方
法で送られるデータや過去に符号化された音声に基づい
て行ったから、このように周期化の程度の制御を行うた
めの情報を特に復号側に伝送する必要がないが、周期化
の程度を示す独立のパラメータを伝送してもよい。何れ
にしても、復号側でも符号化側と全く同様の処理を行
う。また周期化の程度（数）に応じて雑音符号帳１７の
うちのどの符号ベクトルを周期化するかを予め決めてお
く。In addition to estimating the periodicity of the current frame, the pitch periodicity of speech already decoded in the past may be used. That is, the decoded speech is the encoder,
Since it is commonly obtained by the decoders and the pitch periodicity does not change suddenly in successive frames, the periodicity of past frames may be estimated. The estimation of the periodicity of the past frame is performed, for example, from the magnitude of the autocorrelation function. In the above description, the estimation of the periodicity was performed based on the data sent by the conventional encoding method and the speech coded in the past, so that the information for controlling the degree of periodicity is particularly decoded. It is not necessary to transmit to the side, but an independent parameter indicating the degree of periodicization may be transmitted. In any case, the decoding side performs exactly the same processing as the encoding side. In addition, which code vector in the random codebook 17 is to be periodicized is determined in advance according to the degree (number) of periodicization.
【００１６】符号器ではこのような周期化個数の制御の
のち、合成したあとの波形と入力波形との歪が最小とな
る符号ベクトルを決定する。復号器でも同様の周期性の
推定を行って周期化個数を制御して音源信号を作成し
て、線形予測合成フィルタによって最終的出力を得る。
雑音符号帳の符号ベクトルの周期化の程度を制御するの
は周期化する符号ベクトルの数を制御する場合に限ら
ず、１つの符号ベクトルにおける周期化に漏れをもた
せ、その漏れの程度を推定した周期性に応じて制御を行
ってもよい。つまり、この例では漏れ定数γ（０≦γ≦
１）の大きさを推定した周期性の程度に応じて制御す
る。Ｌをピッチ周期、Ａを雑音符号帳１７内の固定の雑
音符号ベクトルとすると漏れをもつ周期化処理された符
号ベクトルＣ′は
Ｃ′（ｉ）＝Ａ（ｉ）
（ｉ≦Ｌの場合）
Ｃ′（ｉ）＝γＣ（ｉ−Ｌ）＋（１−γ）Ａ（ｉ）
（その他の場合）
で表される。すなわちγ＝１で完全に周期化され、γ＝
０で全く周期化されないことになる。例えば図２Ｂに示
すように、γの大、小により周期化符号化ベクトルが変
化する。周期化する符号ベクトルの数を制御する場合は
推定周期性が大きい程数を大とし、漏れγを制御する場
合は推定周期性が大きい程γを小とする。もちろん、以
上２つの実施例、つまり数の制御と漏れγの制御とを組
み合わせた制御も可能である。In the encoder, after controlling the number of periodicizations as described above, the code vector that minimizes the distortion between the waveform after synthesis and the input waveform is determined. The decoder also estimates the periodicity in the same manner and controls the number of periodicizations to create the excitation signal, and obtains the final output by the linear prediction synthesis filter.
Controlling the degree of periodicity of the code vector of the noise codebook is not limited to controlling the number of code vectors to be periodicized, and the periodicity in one code vector is leaked, and the degree of the leak is estimated. The control may be performed according to the periodicity. That is, in this example, the leakage constant γ (0 ≦ γ ≦
The size of 1) is controlled according to the estimated degree of periodicity. If L is a pitch period and A is a fixed noise code vector in the noise codebook 17, a periodic code vector C'having a leak is C '(i) = A (i).
(When i ≦ L) C ′ (i) = γC (i−L) + (1−γ) A (i)
(Other cases) That is, when γ = 1, it is completely periodic, and γ =
When it is 0, it is not cycled at all. For example, as shown in FIG. 2B, the periodic coding vector changes depending on whether γ is large or small. When controlling the number of code vectors to be periodicized, the larger the estimated periodicity is, the larger the number is. When controlling the leakage γ, the larger the estimated periodicity is, the smaller γ is. Of course, the above two embodiments, that is, the control in which the number control and the leakage γ control are combined is also possible.
【００１７】上述で数の制御は雑音符号帳１７の符号ベ
クトルの数の制御のみならず、ＶＳＥＬＰの場合は基本
ベクトルの数の制御であり、漏れγの制御も、基本ベク
トルの周期化の漏れ制御としてもよい。更に上述では、
周期化を当該フレームでの適応符号帳の検索で求めた周
期Ｌで行ったが、前フレームの適応符号帳の検索で得た
周期Ｌ′やＬ／２，２Ｌ，Ｌ′／２，２Ｌ′など適応符
号帳の検索で得た周期と対応したものであればよい。In the above, the control of the number is not only the control of the number of code vectors of the noise codebook 17, but also the control of the number of basic vectors in the case of VSELP, and the control of the leakage γ is also the leakage of the periodicization of the basic vectors. It may be controlled. Furthermore, in the above,
Periodization was performed with the period L obtained by searching the adaptive codebook for the frame, but the period L ′ and L / 2, 2L, L ′ / 2, 2L ′ obtained by searching the adaptive codebook for the previous frame were used. It may be any one corresponding to the cycle obtained by searching the adaptive codebook.
【００１８】[0018]
【発明の効果】以上述べたように、この発明により、ピ
ッチの周期性が高い音声、つまり有声音のフレームでは
最大限にピッチの周期成分を強調するように雑音符号帳
の符号ベクトルを周期化処理し、ピッチの周期性が小さ
い音声、つまり無声音のフレームでは雑音符号帳の符号
ベクトルの周期化処理を行わないようにしており、符号
化音声の歪が減少し、品質が改善される。またこの適応
的処理をすべて既に伝送されている情報や過去の復号化
音声に基づいて行う場合には、従来法と比較して伝送情
報量の増加はない。As described above, according to the present invention, in a speech having a high pitch periodicity, that is, in a voiced frame, the code vector of the random codebook is made periodic so as to maximize the pitch periodic component. By processing the speech, the periodicity of the code vector of the noise codebook is not performed in speech with a small pitch periodicity, that is, in the frame of unvoiced sound, distortion of the encoded speech is reduced, and quality is improved. Further, when all of this adaptive processing is performed based on already transmitted information and past decoded speech, the amount of transmitted information does not increase as compared with the conventional method.
【図１】この発明の実施例の要部を示すブロック図。FIG. 1 is a block diagram showing a main part of an embodiment of the present invention.
【図２】周期化処理で周期性の高い場合と低い場合とを
比較した例を示す図。FIG. 2 is a diagram showing an example in which a case where the periodicity is high and a case where the periodicity is low are compared in the periodical processing.
【図３】線形予測符号化装置の一般的構成を示すブロッ
ク図。FIG. 3 is a block diagram showing a general configuration of a linear predictive coding device.
【図４】ＣＥＬＰにおける雑音符号帳を示す図。FIG. 4 is a diagram showing a random codebook in CELP.
【図５】線形予測符号の復号化装置の一般的構成を示す
ブロック図。FIG. 5 is a block diagram showing a general configuration of a linear predictive code decoding device.
【図６】ＶＳＥＬＰにおける雑音符号帳を示す図。FIG. 6 is a diagram showing a random codebook in VSELP.
【図７】符号ベクトルの周期化を示す図。FIG. 7 is a diagram showing periodicization of a code vector.
Claims (2)
去の駆動音源ベクトルをピッチ周期で繰り返した時系列
ベクトルと、雑音符号帳からの時系列ベクトルとで合成
フィルタを駆動して音声信号を再生することを用いて入
力音声を符号化する音声の励振信号符号化方法におい
て、 上記雑音符号帳の符号ベクトルを、上記ピッチ周期と対
応した周期で繰り返して周期化し、 現在または過去の音声の周期性の程度を推定し、 その推定した周期性の程度に応じて、上記雑音符号帳の
符号ベクトルの周期化の程度を、フレームごとに適応的
に変更することを特徴とする音声の励振信号符号化方
法。1. A speech signal is produced by driving a synthesis filter in a frame unit with a time series vector obtained by repeating a past driving excitation vector from an adaptive codebook in a pitch cycle and a time series vector from a noise codebook. In a method of encoding a voice excitation signal, which encodes an input voice by using reproduction, a code vector of the noise codebook is repeatedly cycled at a period corresponding to the pitch period, and a current or past voice period is obtained. Excitation signal code for speech, characterized in that the degree of periodicity is estimated and the degree of periodicity of the code vector of the noise codebook is adaptively changed for each frame according to the estimated degree of periodicity. Method.
去の駆動音源ベクトルをピッチ周期で繰り返した時系列
ベクトルと、雑音符号帳からの時系列ベクトルとで合成
フィルタを駆動して音声信号を再生する音声信号復号化
方法において、 上記雑音符号帳の符号ベクトルを、上記ピッチ周期と対
応して周期化し、 現在または過去の音声の周期性の程度を推定し、 その推定した周期性の程度に応じて、上記雑音符号帳の
符号ベクトルの周期化の程度をフレーム毎に適応的に変
更することを特徴とする音声信号復号化方法。2. A speech signal is generated by driving a synthesis filter on a frame-by-frame basis with a time-series vector obtained by repeating a past driving excitation vector from the adaptive codebook in a pitch cycle and a time-series vector from a noise codebook. In the audio signal decoding method to be reproduced, the code vector of the random codebook is made periodic corresponding to the pitch cycle, the degree of periodicity of the current or past speech is estimated, and the estimated degree of periodicity is set. According to the method, the degree of periodicity of the code vector of the noise codebook is adaptively changed for each frame.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP3167124A JP2613503B2 (en) | 1991-07-08 | 1991-07-08 | Speech excitation signal encoding / decoding method |
| US07/886,013 US5396576A (en) | 1991-05-22 | 1992-05-20 | Speech coding and decoding methods using adaptive and random code books |
| DE69227401T DE69227401T2 (en) | 1991-05-22 | 1992-05-21 | Method for coding and decoding speech signals |
| EP92108633A EP0514912B1 (en) | 1991-05-22 | 1992-05-21 | Speech coding and decoding methods |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP3167124A JP2613503B2 (en) | 1991-07-08 | 1991-07-08 | Speech excitation signal encoding / decoding method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JPH0519796A true JPH0519796A (en) | 1993-01-29 |
| JP2613503B2 JP2613503B2 (en) | 1997-05-28 |
Family
ID=15843888
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP3167124A Expired - Lifetime JP2613503B2 (en) | 1991-05-22 | 1991-07-08 | Speech excitation signal encoding / decoding method |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP2613503B2 (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH086597A (en) * | 1994-06-21 | 1996-01-12 | Nec Corp | Device and method for coding exciting signal of voice |
| WO2000003385A1 (en) * | 1998-07-13 | 2000-01-20 | Nec Corporation | Voice encoding/decoding device |
| WO2000008633A1 (en) * | 1998-08-06 | 2000-02-17 | Matsushita Electric Industrial Co., Ltd. | Exciting signal generator, voice coder, and voice decoder |
| KR100427752B1 (en) * | 1995-10-26 | 2004-07-19 | 소니 가부시끼 가이샤 | Speech coding method and apparatus |
| JP2014164091A (en) * | 2013-02-25 | 2014-09-08 | Nippon Telegr & Teleph Corp <Ntt> | Encoder, decoder, method of them, program, and recording medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03101800A (en) * | 1989-06-28 | 1991-04-26 | Fujitsu Ltd | Voice encoding system |
-
1991
- 1991-07-08 JP JP3167124A patent/JP2613503B2/en not_active Expired - Lifetime
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03101800A (en) * | 1989-06-28 | 1991-04-26 | Fujitsu Ltd | Voice encoding system |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH086597A (en) * | 1994-06-21 | 1996-01-12 | Nec Corp | Device and method for coding exciting signal of voice |
| KR100427752B1 (en) * | 1995-10-26 | 2004-07-19 | 소니 가부시끼 가이샤 | Speech coding method and apparatus |
| WO2000003385A1 (en) * | 1998-07-13 | 2000-01-20 | Nec Corporation | Voice encoding/decoding device |
| US6856955B1 (en) | 1998-07-13 | 2005-02-15 | Nec Corporation | Voice encoding/decoding device |
| WO2000008633A1 (en) * | 1998-08-06 | 2000-02-17 | Matsushita Electric Industrial Co., Ltd. | Exciting signal generator, voice coder, and voice decoder |
| JP2014164091A (en) * | 2013-02-25 | 2014-09-08 | Nippon Telegr & Teleph Corp <Ntt> | Encoder, decoder, method of them, program, and recording medium |
Also Published As
| Publication number | Publication date |
|---|---|
| JP2613503B2 (en) | 1997-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP3346765B2 (en) | Audio decoding method and audio decoding device | |
| US6427135B1 (en) | Method for encoding speech wherein pitch periods are changed based upon input speech signal | |
| US5602961A (en) | Method and apparatus for speech compression using multi-mode code excited linear predictive coding | |
| JP3180762B2 (en) | Audio encoding device and audio decoding device | |
| WO2001052241A1 (en) | Multi-mode voice encoding device and decoding device | |
| JPH09281998A (en) | Voice coding device | |
| JPH0990995A (en) | Speech coding device | |
| JPH08272395A (en) | Voice encoding device | |
| JP3582589B2 (en) | Speech coding apparatus and speech decoding apparatus | |
| US7024354B2 (en) | Speech decoder capable of decoding background noise signal with high quality | |
| JP3275247B2 (en) | Audio encoding / decoding method | |
| JPH0944195A (en) | Voice encoding device | |
| JP2613503B2 (en) | Speech excitation signal encoding / decoding method | |
| JP2538450B2 (en) | Speech excitation signal encoding / decoding method | |
| JPH08328597A (en) | Sound encoding device | |
| JP3144284B2 (en) | Audio coding device | |
| JP3299099B2 (en) | Audio coding device | |
| EP1154407A2 (en) | Position information encoding in a multipulse speech coder | |
| JP3319396B2 (en) | Speech encoder and speech encoder / decoder | |
| JP3153075B2 (en) | Audio coding device | |
| JP3232701B2 (en) | Audio coding method | |
| JP3099836B2 (en) | Excitation period encoding method for speech | |
| JP3292227B2 (en) | Code-excited linear predictive speech coding method and decoding method thereof | |
| JP2002073097A (en) | Celp type voice coding device and celp type voice decoding device as well as voice encoding method and voice decoding method | |
| JPH08185199A (en) | Voice coding device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20090227 Year of fee payment: 12 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20090227 Year of fee payment: 12 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20100227 Year of fee payment: 13 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20110227 Year of fee payment: 14 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20110227 Year of fee payment: 14 |
|
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20120227 Year of fee payment: 15 |
|
| EXPY | Cancellation because of completion of term | ||
| FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20120227 Year of fee payment: 15 |