JP6133454B2 - Audio signal processing method and audio signal processing apparatus - Google Patents

Description

  The present invention relates to an audio signal processing method and an audio signal processing apparatus for processing an audio signal.

  When an audio signal is encoded and packetized and transmitted over the Internet network like an IP (Internet Protocol) telephone, the packet may be lost due to network congestion or the like (hereinafter this phenomenon is referred to as “packet”). "Loss"). When a packet loss occurs, a necessary voice code is lost, so that voice decoding cannot be performed and sound interruption occurs. There is a voice packet loss concealment technique as a technique for preventing sound interruption caused by packet loss. The voice packet loss concealment technique detects a packet loss and generates a pseudo voice signal (hereinafter, “concealment signal”) corresponding to the lost packet.

  When using speech encoding while updating the internal state of the encoder / decoder as the speech encoding method, the encoding parameters that should be received cannot be obtained. The concealment technique also updates the internal state of the decoder with pseudo-generated parameters.

CELP (Code Excited Linear Prediction) coding is widely used as a method for performing speech coding while updating the internal state of the encoder / decoder. In CELP coding, an autoregressive model is assumed, and a speech signal is synthesized by filtering the excitation signal e (n) with an all-pole synthesis filter a (i). That is, the audio signal s (n) is synthesized according to the following equation. a (i) is a linear prediction coefficient (LP (Linear Prediction) coefficient), and a value such as P = 16 is used as the order.

  CELP coding has an ISF (Immittance Spectral Frequency) parameter that expresses a linear prediction coefficient mathematically equivalently and a past excitation signal as an internal state. When packet loss occurs, these are generated in a pseudo manner, and there is a divergence from the parameters that were originally obtained by decoding. Synthetic speech inconsistency caused by parameter divergence is perceived as noise by the listener, greatly impairing subjective quality.

  In the following, the configuration and operation of a speech decoder that performs speech packet loss concealment will be described, taking as an example the case where CELP coding is used as the speech coding method.

  FIG. 1 and FIG. 2 show the configuration and operation of the speech decoder. As shown in FIG. 1, the speech decoder 1 includes a packet loss detection unit 11, a speech code decoding unit 12, a concealment signal generation unit 13, and an internal state buffer 14.

  When the packet loss detection unit 11 normally receives the voice packet, the packet loss detection unit 11 sends the control signal and the voice code included in the voice packet to the voice code decoding unit 12 (normal reception: YES in step S100 of FIG. 2) ). Thereafter, the speech code decoding unit 12 performs speech code decoding and internal state update as described later (steps S200 and S400 in FIG. 2). On the other hand, when the voice packet cannot be normally received, the packet loss detection unit 11 sends a control signal to the concealment signal generation unit 13 (packet loss: NO in step S100 in FIG. 2). Thereafter, the concealment signal generation unit 13 generates a concealment signal and updates the internal state as described later (steps S300 and S400 in FIG. 2). The processes in steps S100 to S400 in FIG. 2 are repeated until the end of communication (until YES is determined in step S500).

、符号化された第一から第四サブフレームのピッチラグT^j _p、第一から第四サブフレームの符号化された適応符号帳ゲインg^j _p、第一から第四サブフレームの符号化された固定符号帳ゲインg^j _c、第一から第四サブフレームの符号化された固定符号帳ベクトルc^j(n)を含む。ISFパラメータの代わりに数学的に等価な表現であるLSF（line spectral frequency）パラメータを用いてもよい。以下の議論では、ISFパラメータを用いた説明を行うが、LSFパラメータを用いた場合も同じ議論が成り立つ。 Speech code is at least encoded ISF parameter

, Encoded first to fourth subframe pitch lag T ^j _p , first to fourth subframe encoded adaptive codebook gain g ^j _p , first to fourth subframe encoded It includes a fixed codebook gain g ^j _c and encoded fixed codebook vectors c ^j (n) of the first to fourth subframes. An LSF (line spectral frequency) parameter that is a mathematically equivalent expression may be used instead of the ISF parameter. In the following discussion, explanation will be made using ISF parameters, but the same argument holds when using LSF parameters.

及び

の等価表現であるISP（Immittance Spectral Pair）パラメータ

、ISF残差パラメータ

、過去のピッチラグT^j _p、過去の適応符号帳ゲインg^j _p、過去の固定符号帳ゲインg^j _c、適応符号帳u(n)を含む。それぞれ過去何サブフレーム分のパラメータを含むかは設計方針による。本明細書では、１フレームが４サブフレームを含むことを仮定するが、設計方針により他の値としてもよい。 The internal state buffer contains past ISF parameters

as well as

ISP (Immittance Spectral Pair) parameter, which is the equivalent expression of

, ISF residual parameters

, Past pitch lag T ^j _p , past adaptive codebook gain g ^j _p , past fixed codebook gain g ^j _c , and adaptive codebook u (n). How many past subframe parameters are included depends on the design policy. In this specification, it is assumed that one frame includes 4 subframes, but other values may be used depending on the design policy.

<For normal reception>
FIG. 3 shows a functional configuration example of the speech code decoding unit 12. As shown in FIG. 3, the speech code decoder 12 includes an ISF decoder 120, a stability processor 121, an LP coefficient calculator 122, an adaptive codebook calculator 123, a fixed codebook decoder 124, and a gain decoder 125. , An excitation vector synthesis unit 126, a post filter 127, and a synthesis filter 128. However, the post filter 127 is not an essential component. In FIG. 3, for convenience of explanation, the internal state buffer 14 is indicated by a two-dot chain line in the speech code decoding unit 12. However, the internal state buffer 14 is not included in the speech code decoding unit 12. The internal state buffer 14 shown in FIG. The same applies to the configuration diagrams of the subsequent speech code decoding units.

  FIG. 4 shows a configuration diagram of the LP coefficient calculation unit 122, and FIG. 5 shows a processing flow for calculating the LP coefficient from the encoded ISF parameter. As shown in FIG. 4, the LP coefficient calculation unit 122 includes an ISF-ISP conversion unit 122A, an ISP interpolation unit 122B, and an ISP-LPC conversion unit 122C.

  First, the functional configuration and operation related to the process (FIG. 5) for calculating the LP coefficient from the encoded ISF parameter will be described.

を求め、ISFパラメータ

を次式に従い算出する（図５のステップＳ１）。ここで、mean_iは、事前に学習等で求めた平均ベクトルである。

The ISF decoding unit 120 decodes the encoded ISF parameter to generate an ISF residual parameter

ISF parameter

Is calculated according to the following equation (step S1 in FIG. 5). Here, mean _i is an average vector obtained by learning or the like in advance.

、AR予測の重み係数をρ_iとした。

In addition, although the example which uses MA prediction for calculation of ISF parameter was described here, it is good also as a structure which calculates ISF parameter using AR prediction as follows. Here, the ISF parameter of the previous frame

The weight coefficient for AR prediction is ρ _i .

In order to ensure the stability of the filter, the stability processing unit 121 performs processing according to the following equation so as to leave an interval of 50 Hz or more between the elements of the ISF parameter (step S2 in FIG. 5). The ISF parameter expresses the shape of the speech spectrum envelope with a line spectrum, and as the distance from each other decreases, the spectrum peak increases and resonance occurs. Therefore, it is necessary to ensure stability so that the gain at the peak of the spectrum does not become too large. Here, min_dist is the minimum ISF interval, and isf_min is the minimum value of ISF necessary to secure the min_dist interval. isf_min is sequentially updated by adding the min_dist interval to the value of the adjacent ISF. On the other hand, isf_max is the maximum value of the ISF necessary to secure the min_dist interval. isf_max is sequentially updated by subtracting the min_dist interval from the value of the adjacent ISF.

をISPパラメータ

に変換する（図５のステップＳ３）。ここで、Cは事前に定めた定数である。

The ISF-ISP converter 122A in the LP coefficient calculator 122 follows the following formula:

ISP parameters

(Step S3 in FIG. 5). Here, C is a predetermined constant.

と、上記ISPパラメータ

から、以下の式に従いサブフレーム毎のISPパラメータを算出する（図５のステップＳ４）。補間にあたっては、別の係数を用いてもよい。

The ISP interpolation unit 122B uses the past ISP parameters included in the internal state buffer 14.

And the above ISP parameters

From this, ISP parameters for each subframe are calculated according to the following equation (step S4 in FIG. 5). In the interpolation, another coefficient may be used.

に変換する（図５のステップＳ５）。具体的変換手順として、非特許文献１に記載の処理手順を用いることができる。ここで、先読み信号に含まれるサブフレームの数を４としたが、サブフレームの数は設計方針により変更してもよい。 ISP-LPC converter 122C converts ISP parameters for each subframe into LP coefficients.

(Step S5 in FIG. 5). As a specific conversion procedure, the processing procedure described in Non-Patent Document 1 can be used. Here, although the number of subframes included in the prefetch signal is four, the number of subframes may be changed according to the design policy.

  Next, other configurations and operations in the speech code decoding unit 12 will be described.

The adaptive codebook calculating unit 123 decodes the encoded pitch lag and calculates the pitch lag T ^j _p of the first to fourth subframes. Next, adaptive codebook calculating section 123 calculates an adaptive codebook vector for each subframe according to the following equation using adaptive codebook u (n). The adaptive codebook vector is calculated by interpolating the adaptive codebook u (n) with the FIR filter Int (i). Here, the length of the adaptive codebook is N _adapt . The filter Int (i) used for the interpolation is a FIR filter having a predetermined length 2l + 1, and L ′ is the number of samples in the subframe. By using the interpolation filter Int (i), the pitch lag can be used up to the precision after the decimal point. For details of the interpolation filter, the method described in Non-Patent Document 1 can be used.

Fixed codebook decoding section 124 decodes the encoded fixed codebook vector to obtain fixed codebook vectors c ^j (n) for the first to fourth subframes.

  The gain decoding unit 125 decodes the encoded adaptive codebook gain and the encoded fixed codebook gain, and acquires the adaptive codebook gain and the fixed codebook gain of the first to fourth subframes. For example, the adaptive codebook gain and the fixed codebook gain can be decoded by the following method described in Non-Patent Document 1. According to the following method described in Non-Patent Document 1, since the inter-frame prediction is not used like the AMR-WB gain encoding, the packet loss tolerance can be enhanced.

  For example, the gain decoding unit 125 acquires a fixed codebook gain according to the following processing flow.

First, gain decoding section 125 calculates the power of the fixed codebook vector. Here, let N _{s be} the length of the subframe.

と量子化固定符号帳ゲイン

を得る。量子化固定符号帳ゲインと上記固定符号帳ベクトルのパワーから、以下の通り予測固定符号帳ゲインを算出する。

Next, gain decoding section 125 decodes the vector-quantized gain parameter, and performs adaptive codebook gain.

And quantized fixed codebook gain

Get. From the quantized fixed codebook gain and the power of the fixed codebook vector, the predicted fixed codebook gain is calculated as follows.

を復号し、予測ゲインに乗算することにより、固定符号帳ゲインを得る。

Finally, the gain decoding unit 125 performs the prediction coefficient

And a prediction gain is multiplied to obtain a fixed codebook gain.

The excitation vector synthesizer 126 multiplies the adaptive codebook vector by the adaptive codebook gain, multiplies the fixed codebook vector by the fixed codebook gain, and obtains the sum of them by the following equation. To get.

  The post filter 127 performs post-processing such as pitch enhancement, noise enhancement, and low frequency enhancement on the excitation signal vector. For pitch emphasis, noise emphasis, and low frequency emphasis, the method described in Non-Patent Document 1 can be used.

The synthesis filter 128 synthesizes a decoded signal using the excitation signal as a driving sound source by linear predictive inverse filtering.

In the encoder, when pre-emphasis is performed, de-emphasis is performed.

  On the other hand, when pre-emphasis is not performed in the encoder, de-emphasis is not performed.

  Hereinafter, operations related to the internal state update will be described.

The LP coefficient calculation unit 122 updates the internal state of the ISF parameter with a vector calculated by the following equation for parameter interpolation at the time of packet loss.

Here, ω _i ^(−j) is the ISF parameter before j frames stored in the buffer. ω _i ^C is an ISF parameter in the utterance section obtained by learning or the like in advance. β is a constant and can be a value such as 0.75, but is not limited thereto. ω _i ^C and β may be changed by an index representing the property of the encoding target frame, for example, as in the ISF concealment described in Non-Patent Document 1.

Furthermore, the LP coefficient calculation unit 122 also updates the internal state of the ISF residual parameter according to the following equation.

The excitation vector combining unit 126 updates the internal state with the excitation signal vector according to the following equation.

Furthermore, the excitation vector synthesis unit 126 updates the internal state of the gain parameter using the following equation.

なお、ここでは、(−２≦ｊ＜Ｍ_ｌａ)とするが、ｊの範囲として設計方針により異なる値を選択してもよい。 Adaptive codebook calculating section 123 updates the internal state of the pitch lag parameter according to the following equation.

Here, (−2 ≦ j <M _la ) is assumed, but a different value may be selected as the range of j depending on the design policy.

<In case of packet loss>
FIG. 6 shows a functional configuration example of the concealment signal generation unit 13. As shown in FIG. 6, the concealment signal generation unit 13 includes an LP coefficient interpolation unit 130, a pitch lag interpolation unit 131, a gain interpolation unit 132, a noise signal generation unit 133, a post filter 134, a synthesis filter 135, and an adaptive codebook calculation unit. 136 and an excitation vector synthesis unit 137. However, the post filter 134 is not an essential component.

を次式により算出する。なお、ω_i ^(-j)は、バッファに格納されたｊフレーム前のISFパラメータである。

ここで、

はパケットを正常に受信した際に算出されたISFパラメータの内部状態である。αも定数であり、0.9のような値とすることができるが、これに限定されない。αは、例えば非特許文献１記載のISFコンシールメントのように、符号化対象フレームの性質を表すインデクスにより変化させてもよい。 The LP coefficient interpolation unit 130

Is calculated by the following equation. Note that ω _i ^(−j) is an ISF parameter before j frames stored in the buffer.

here,

Is an internal state of the ISF parameter calculated when the packet is normally received. α is also a constant and can be a value such as 0.9, but is not limited thereto. α may be changed by an index representing the property of the encoding target frame, for example, as in the ISF concealment described in Non-Patent Document 1.

  The procedure for obtaining the LP coefficient from the ISF parameter is the same as that when the packet is normally received.

を用いて、ピッチラグの予測値

を算出する。具体的な処理手順として、非特許文献１の手法を用いることができる。 The pitch lag interpolation unit 131 is an internal state parameter related to pitch lag.

Is used to predict pitch lag.

Is calculated. As a specific processing procedure, the technique of Non-Patent Document 1 can be used.

The gain interpolation unit 132 can use a method according to the following equation described in Non-Patent Document 1 for interpolation of the fixed codebook gain.

  The noise signal generation unit 133 generates white noise having the same length as the fixed codebook vector, and substitutes it as the fixed codebook vector.

  The operations of the post filter 134, the synthesis filter 135, the adaptive codebook calculation unit 136, and the excitation vector synthesis unit 137 are the same as the operations performed when the packet is normally received.

The internal state update is the same as when the packet is normally received except for the ISF residual parameter. The ISF residual parameter is updated by the LP coefficient interpolation unit 130 according to the following equation.

International Publication No. 2002/035520 International Publication 2008/108080

ITU-T Recommendation G.718, June 2008

  As described above, since CELP encoding has an internal state, sound quality is deteriorated due to a difference between a parameter obtained by interpolation at the time of packet loss and a parameter that should be originally used for decoding. In particular, since ISF parameters are subjected to predictive coding within and between frames, there is a problem that the influence of packet loss continues even after the packet loss is recovered.

  More specifically, a problem has been confirmed in which the power suddenly increases in the first frame recovered from the packet loss that occurred near the start of the voice. This is because the impulse response of the LP coefficient calculated from the ISF coefficient obtained by the interpolation process at the time of packet loss is higher than that originally assumed by the decoder at the beginning of the voice where the power of the excitation signal is high. Is perceived as a discontinuous sound that is unpleasant in subjective quality.

  The method of Patent Document 1 generates an ISF coefficient interpolated with respect to a lost frame, but for the first frame restored from the loss, an ISF parameter is generated by normal decoding. It cannot be suppressed.

  On the other hand, in the method of Patent Document 2, a gain adjustment parameter (normalized prediction residual power) obtained on the encoding side is transmitted and used for power adjustment on the decoding side, so that the excitation signal of the frame lost frame is lost. Power can be suppressed and a sudden increase in power can be prevented.

  FIG. 7 shows a functional configuration example of the speech decoder 1X corresponding to the technique of Patent Document 2, and FIG. 8 shows a functional configuration example of the concealment signal generation unit 13X. In Patent Document 2, a voice packet includes at least auxiliary information of normalized prediction residual power in addition to the parameters described in the conventional method.

  The normalized prediction residual power decoding unit 15 included in the audio signal generation unit 1X decodes the auxiliary information of the normalized prediction residual power from the audio packet, calculates the reference normalized prediction residual power, and the concealment signal generation unit Output to 13X.

  Among the constituent elements of the concealment signal generation unit 13X, the constituent elements other than the normalized prediction residual adjustment unit 138 are the same as those in the conventional technique described above, and therefore only the normalized prediction residual adjustment unit 138 will be described below.

  The normalized prediction residual adjustment unit 138 calculates normalized prediction residual power from the LP coefficient output from the LP coefficient interpolation unit 130. Next, the normalized prediction residual adjustment unit 138 calculates a synthesis filter gain adjustment coefficient using the normalized prediction residual power and the reference normalized prediction residual power. Finally, the normalized prediction residual adjustment unit 138 multiplies the excitation signal by the synthesis filter gain adjustment coefficient and outputs the result to the synthesis filter 135.

  According to the technique of Patent Document 2 described above, the power of the concealment signal at the time of packet loss can be suppressed in the same way as at the time of normal reception. However, in low bit rate speech coding, it is necessary to transmit the gain adjustment parameter. It is difficult to ensure a high bit rate. In addition, since the process is performed by the concealment signal generation unit, it is difficult to cope with a sudden change in power caused by the mismatch of the ISF parameters in the recovery frame.

  Therefore, an object of the present invention is to reduce the discontinuous sound that may occur when returning from a packet loss at the time of starting voice and improve the subjective quality.

  An audio signal processing method according to an embodiment of the present invention is an audio signal processing method executed by an audio signal processing apparatus, which is auxiliary information transmitted from an encoder and obtained as a result of audio packet decoding A decoding step for decoding auxiliary information related to speech discontinuity, an estimation step for estimating discontinuity of decoded speech using the auxiliary information decoded by the decoding step, and an estimation step for estimating discontinuity. A discontinuity of the decoded speech, and a correction step of correcting the discontinuity of the decoded speech when the estimation step does not estimate that the discontinuity is not estimated. This occurs when the amplitude of the decoded speech obtained as a result of speech packet decoding suddenly increases with respect to the speech packet normally received first after the occurrence. That.

  The audio signal processing method described above further includes a detection step of detecting a packet loss of the audio packet transmitted from the encoder, and the correction step is not performed by the estimation step when the packet loss is detected by the detection step. If it is estimated to be continuous, the discontinuity of the decoded speech may be corrected.

  An audio signal processing apparatus according to an embodiment of the present invention includes an auxiliary information decoder that decodes auxiliary information related to discontinuity of decoded audio obtained as a result of audio packet decoding, which is auxiliary information transmitted from an encoder. A discontinuity estimator that estimates the discontinuity of the decoded speech using the auxiliary information decoded by the auxiliary information decoder, and a discontinuity of the decoded speech that is estimated to be discontinuous by the discontinuity estimator A discontinuity corrector that does not correct the discontinuity of the decoded speech if it is not estimated to be discontinuous by the discontinuity estimator, the discontinuity being the first after a packet loss has occurred This is because the amplitude of the decoded speech obtained as a result of speech packet decoding suddenly increases with respect to the speech packet normally received normally.

  According to the present invention as described above, it is possible to reduce the discontinuous sound that may occur when returning from a packet loss at the time of voice start, and improve subjective quality.

It is a block diagram of a speech decoder. It is a processing flow of a speech decoder. It is a functional block diagram of a speech coder / decoder. It is a functional block diagram of a LP coefficient calculation part. It is a processing flow which calculates LP coefficient. It is a functional block diagram of a concealment signal generation part. FIG. 10 is a configuration diagram of a speech decoder of Patent Document 2. FIG. 10 is a functional configuration diagram of a concealment signal generation unit of Patent Document 2. It is a function block diagram of the audio | voice code | cord decoding part in 1st Embodiment. It is a processing flow of the LP coefficient calculation part in 1st Embodiment. It is a function block diagram of the audio | voice code | cord decoding part in 1st Embodiment. It is a processing flow of the 2nd stability treating part in a 1st embodiment modification 1. It is a function block diagram of the audio | voice code decoding part in 2nd Embodiment. It is a functional block diagram of the LP coefficient calculation part in 2nd Embodiment. It is a processing flow of LP coefficient calculation in the second embodiment. It is a block diagram of the audio | voice encoder in 4th Embodiment. It is a block diagram of the audio | voice encoder in 4th Embodiment. It is a block diagram of the LP analyzer / encoder in the fourth embodiment. It is a processing flow of the LP analyzer / encoder in the fourth embodiment. It is a function block diagram of the audio | voice code | cord decoding part in 4th Embodiment. It is a processing flow of the LP coefficient calculation part in 4th Embodiment. It is a block diagram of the LP analyzer / encoder in the fifth embodiment. It is a processing flow of the LP analyzer / encoder in the fifth embodiment. It is a function block diagram of the audio | voice code | cord decoding part in 4th Embodiment. It is a processing flow of the LP coefficient calculation part in 5th Embodiment. It is a block diagram of the audio | voice decoder in 7th Embodiment. It is a processing flow of the audio | voice decoder in 7th Embodiment. It is a function block diagram of the audio | voice code decoding part in 7th Embodiment. It is a processing flow of LP coefficient calculation in the seventh embodiment. It is a figure which shows the hardware structural example of a computer. It is an external view of a computer. (A), (b), (c) and (d) is a figure which shows the various structural examples of an audio | voice signal processing program.

  Hereinafter, preferred embodiments of an audio signal processing device, an audio signal processing method, and an audio signal processing program according to the present invention will be described in detail with reference to the drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.

[First Embodiment]
Since the speech signal processing apparatus according to the first embodiment has the same configuration as the speech decoder 1 of FIG. 1 described above and is characterized by the speech code decoding unit, the speech code decoding unit will be described below.

  FIG. 9 is a functional configuration diagram of the speech code decoding unit 12A in the first embodiment, and FIG. 10 is a flowchart of LP coefficient calculation processing. The speech code decoding unit 12A in FIG. 9 has a discontinuity detection unit 129 added to the configuration in FIG. 3 described above. Since the difference from the prior art is only the LP coefficient calculation process, the operation of each part related to the LP coefficient calculation process will be described here.

The discontinuity detecting unit 129 refers to the fixed codebook gain g _c ⁰ obtained by decoding and the fixed codebook gain g _c ⁻¹ included in the internal state, and compares the gain change with the threshold according to the following equation: (Step S11 in FIG. 10).

  When the gain change exceeds the threshold, the occurrence of discontinuity is detected (hereinafter also simply referred to as “discontinuity detection”), and a control signal related to the detection result of the discontinuity occurrence is output to the stability processing unit 121.

Note that the following equation may be used in comparing the gain change and the threshold.

Further, the maximum fixed codebook gain of the first to fourth subframes included in the current frame is g _c ^(c) , and the minimum fixed codebook gain included in the internal state is g _c ^{(p )} , A change in gain and a threshold value may be compared by the following equation.

Of course, the following equation may be used.

In the above example of the first embodiment, the fixed codebook gain g _c ⁻¹ of the fourth subframe of the immediately preceding frame (lost frame) and the fixed codebook gain g _c ⁰ of the first subframe of the current frame are used. Although an example of performing discontinuity detection has been shown, after calculating an average value for each of the fixed codebook gain included in the internal state and the fixed codebook gain included in the current frame, the gain change is compared with the threshold value. You may make it perform.

  The ISF decoding unit 120 performs the same operation as in the conventional technique (step S12 in FIG. 10).

  When the discontinuity detecting unit 129 detects discontinuity, the stability processing unit 121 corrects the ISF parameter by the following process (step S13 in FIG. 10).

について、各要素間に通常よりＭ_-1倍の間隔をあける処理を行う。通常よりも非常に大きな間隔を与えることによりスペクトル包絡における過大なピークとディップを抑制する効果を与える。ここで、min_distは最小のISF間隔であり、isf_minは、min_distの間隔を確保するために必要なISFの最小値である。isf_minは、隣のISFの値にmin_distの間隔を加算することにより順次更新を行う。一方、isf_maxは、min_distの間隔を確保するために必要なISFの最大値である。isf_maxは、隣のISFの値からmin_distの間隔を減算することにより順次更新を行う。

First, the stability processing unit 121 sets the ISF parameter stored in the internal state buffer 14.

With respect to, a process is performed in which each element is spaced M ₋₁ times larger than usual. By giving a much larger interval than usual, an effect of suppressing an excessive peak and dip in the spectrum envelope is given. Here, min_dist is the minimum ISF interval, and isf_min is the minimum value of ISF necessary to secure the min_dist interval. isf_min is sequentially updated by adding the min_dist interval to the value of the adjacent ISF. On the other hand, isf_max is the maximum value of the ISF necessary to secure the min_dist interval. isf_max is sequentially updated by subtracting the min_dist interval from the value of the adjacent ISF.

Next, the stability processing unit 121 performs a process of setting an interval M ₀ times larger than usual between the elements for the ISF parameter of the current frame. Here, 1 <M ₀ <M ₋₁ , but either M ₋₁ or M ₀ may be set to 1, and the other may be set to a value larger than 1.

In addition, when the discontinuity detector does not detect discontinuity, the stability processing unit 121 performs the following processing in the same manner as in the normal decoding process.

  The minimum interval between elements when discontinuity is detected may be changed according to the frequency of the ISF. The minimum interval between elements when discontinuity is detected may be different from the minimum interval between elements in a normal decoding process.

をそれぞれ、ISPパラメータ

に変換する（図１０のステップＳ１４）。ここで、Cは事前に定めた定数である。

The ISF-ISP conversion unit 122A in the LP coefficient calculation unit 122 follows the following formula and sets the ISF parameter

Respectively, ISP parameters

(Step S14 in FIG. 10). Here, C is a predetermined constant.

と、上記ISPパラメータ

から、以下の式に従い、サブフレーム毎のISPパラメータを算出する（図１０のステップＳ１５）。補間にあたっては、別の係数を用いてもよい。

ISP interpolator 122B uses the past ISP parameters

And the above ISP parameters

Then, ISP parameters for each subframe are calculated according to the following formula (step S15 in FIG. 10). In the interpolation, another coefficient may be used.

に変換する（図１０のステップＳ１６）。ここで、先読み信号に含まれるサブフレームの数を４としたが、サブフレームの数は設計方針により変更してもよい。具体的変換手順として、非特許文献１に記載の処理手順を用いることができる。 ISP-LPC converter 122C converts ISP parameters for each subframe into LP coefficients.

(Step S16 in FIG. 10). Here, although the number of subframes included in the prefetch signal is four, the number of subframes may be changed according to the design policy. As a specific conversion procedure, the processing procedure described in Non-Patent Document 1 can be used.

を次式に従い更新する。

このとき、不連続が検出された場合でも、ISF-ISP変換部１２２Ａは、以下の手順を実施することにより、ISFパラメータの算出結果を用いて、内部状態バッファに記憶されたISFパラメータ

を更新してもよい。

In addition, the ISF-ISP conversion unit 122A uses the ISF parameters stored in the internal state buffer 14.

Is updated according to the following equation.

At this time, even if discontinuity is detected, the ISF-ISP conversion unit 122A performs the following procedure to use the ISF parameter calculation result to store the ISF parameter stored in the internal state buffer.

May be updated.

  As in the first embodiment described above, the discontinuity of the decoded speech can be estimated by the quantized codebook gain used for the excitation signal calculation, and the ISF is determined according to the discontinuity estimation result. The / LSF parameter (for example, the interval between each element of the ISF / LSF parameter given to guarantee the stability of the synthesis filter) can be modified. Thereby, the discontinuous sound which may occur when returning from the packet loss at the time of starting voice can be reduced, and the subjective quality can be improved.

[Modification of First Embodiment]
FIG. 11 shows a functional configuration diagram of a speech code decoding unit 12S according to a modification of the first embodiment. Since the only difference from the configuration of the prior art in FIG. 3 is the discontinuity detection unit 129 and the second stability processing unit 121S, these operations will be described. The second stability processing unit 121S includes a gain adjustment unit 121X and a gain multiplication unit 121Y. The processing flow of the second stability processing unit 121S is shown in FIG.

The discontinuity detection unit 129 uses the fixed codebook gain g _c ⁰ obtained by decoding and the fixed codebook gain g _c ⁻¹ included in the internal state in the same manner as the discontinuity detection unit 129 of the first embodiment. Reference is made to compare the gain change with the threshold. Then, the discontinuity detection unit 129 sends a control signal including information regarding whether or not the gain change exceeds the threshold value to the gain adjustment unit 121X.

The gain adjustment unit 121X reads information about whether or not the gain change exceeds the threshold value from the control signal, and outputs the predetermined gain g _on to the gain multiplication unit 121Y when the gain change exceeds the threshold value. To do. On the other hand, when the gain change does not exceed the threshold value, the gain adjustment unit 121X outputs a predetermined gain g _off to the gain multiplication unit 121Y. Such an operation of the gain adjusting unit 121X corresponds to step S18 in FIG.

The gain multiplication unit 121Y multiplies the synthesized signal output from the synthesis filter 128 by the gain g _on or the gain g _off (step S19 in FIG. 12), and outputs the obtained decoded signal.

  Here, an LP coefficient or an ISF parameter is output from the LP coefficient calculation unit 122 and input to the second stability processing unit 121S (indicated by a dotted line from the LP coefficient calculation unit 122 to the gain adjustment unit 121X in FIG. 11). (Configuration). In this case, the gain to be multiplied is determined using the LP coefficient or ISF parameter calculated by the LP coefficient calculation unit 122.

  As in the above modification, the second stability processing unit 121S is added to the speech code decoding unit 12S, and the gain is adjusted according to whether or not the change in gain exceeds the threshold value. Can be obtained.

  The second stability processing unit 121S may multiply the excitation signal by the calculated gain and output it to the synthesis filter 128.

[Second Embodiment]
Since the speech signal processing apparatus according to the second embodiment has the same configuration as the speech decoder 1 of FIG. 1 described above and is characterized by the speech code decoding unit, the speech code decoding unit will be described below. FIG. 13 shows a functional configuration example of the speech code decoder 12B, FIG. 14 shows a functional configuration example related to the LP coefficient calculation processing, and FIG. 15 shows a flow of the LP coefficient calculation processing. The speech code decoding unit 12B of FIG. 13 has a discontinuity detection unit 129 added to the configuration of FIG. 3 described above.

  The ISF decoding unit 120 calculates ISF parameters in the same manner as in the conventional technique (step S21 in FIG. 15).

の各要素間に50Hz以上の間隔をあけるよう処理を行う（図１５のステップＳ２２）。 In the same manner as the conventional technique, the stability processing unit 121 uses the ISF parameter to ensure the stability of the filter.

Processing is performed so as to leave an interval of 50 Hz or more between each element (step S22 in FIG. 15).

  The ISF-ISP conversion unit 122A converts the ISF parameter output from the stability processing unit 121 into an ISP parameter in the same manner as in the first embodiment (step S23 in FIG. 15).

と、ISF-ISP変換部１２２Ａによる変換で得られたISPパラメータ

から、サブフレーム毎のISPパラメータを算出する（図１５のステップＳ２４）。 The ISP interpolator 122B performs past ISP parameters in the same manner as in the first embodiment.

And ISP parameters obtained by conversion by ISF-ISP converter 122A

From this, ISP parameters for each subframe are calculated (step S24 in FIG. 15).

に変換する（図１５のステップＳ２５）。ここで、先読み信号に含まれるサブフレームの数を４としたが、サブフレームの数は設計方針により変更してもよい。 The ISP-LPC converter 122C converts the ISP parameters for each subframe into LP coefficients in the same manner as in the first embodiment.

(Step S25 in FIG. 15). Here, although the number of subframes included in the prefetch signal is four, the number of subframes may be changed according to the design policy.

  The internal state buffer 14 updates the ISF parameter stored in the past with the new ISF parameter.

The discontinuity detection unit 129 reads the LP coefficient of the fourth subframe in the packet lost frame from the internal state buffer 14 and calculates the impulse response power of the LP coefficient of the fourth subframe in the packet lost frame. As the LP coefficient of the fourth subframe in the packet loss frame, the coefficient output from the LP coefficient interpolation unit 130 included in the concealment signal generation unit 13 of FIG. .

Then, the discontinuity detection unit 129 detects discontinuity by the following equation, for example (step S26 in FIG. 15).

  If the gain change does not exceed the threshold value (NO in step S27 in FIG. 15), the discontinuity detection unit 129 does not detect the occurrence of discontinuity, and outputs the LP coefficient from the ISP-LPC conversion unit 122C for processing. finish. On the other hand, when the gain change exceeds the threshold value (YES in step S27 of FIG. 15), the discontinuity detection unit 129 detects the occurrence of discontinuity and performs a stability process on the control signal related to the detection result of the discontinuity occurrence. Send to section 121. When the control signal is received, the stability processing unit 121 corrects the ISF parameter in the same manner as in the first embodiment (step S28 in FIG. 15). Hereinafter, the operations of the ISF-ISP conversion unit 122A, ISP interpolation unit 122B, and ISP-LPC conversion unit 122C (steps S29, S2A, and S2B in FIG. 15) are the same as described above.

  As in the second embodiment described above, the discontinuity of the decoded speech can be estimated by the power of the excitation signal, and as in the first embodiment, the discontinuous sound can be reduced and the subjective quality can be improved. it can.

[Third embodiment]
When the discontinuity is detected, the ISF parameter may be corrected by another method. Since only the stability processing unit 121 is different from the first embodiment in the third embodiment, only the operation of the stability processing unit 121 will be described.

  When the discontinuity detection unit 129 detects discontinuity, the stability processing unit 121 performs the following process to correct the ISF parameter.

について、安定性処理部１２１は、低次P’次元（0＜P’≦P）までのISFパラメータを次式により置き換える。ここで、

とする。

ISF parameters stored in internal state buffer 14

, The stability processing unit 121 replaces the ISF parameters up to the low-order P ′ dimension (0 <P ′ ≦ P) with the following expression. here,

And

Further, the stability processing unit 121 may overwrite the low-order P′-dimensional ISF parameter with the P′-dimensional vector obtained by learning in advance as follows.

とする。

Next, with respect to the ISF parameters of the current frame, the stability processing unit 121 may perform a process of spacing M ₀ times more than usual between the elements as in the first embodiment, or may be determined according to the following equation: May be. here,

And

In addition, the stability processing unit 121 may overwrite the P′-dimensional vector learned in advance.

としてもよい。ただし、復号開始時のフレームにおいてはω_i ^-1を予め決められたP’次元ベクトルω_i ^initとしてもよい。 Furthermore, the P′-dimensional vector may be learned in the decoding process, for example,

It is good. However, in a frame at the start of decoding, ω _i ⁻¹ may be a predetermined ^P′- dimensional vector ω _i ^init .

  The internal state buffer 14 updates the ISF parameter stored in the past with the new ISF parameter.

  As in the third embodiment described above, the interval obtained by equally dividing the ISF / LSF parameter up to a predetermined dimension is the interval between each element of the ISF / LSF parameter that is given to guarantee the stability of the synthesis filter. As in the first and second embodiments, it is possible to reduce discontinuous sounds and improve subjective quality.

[Fourth Embodiment]
In the fourth embodiment, the encoding side detects the occurrence of discontinuity, includes the discontinuity determination code (the code indicating the detection result) in the speech code, and transmits it to the decoding side. The decoding side is included in the speech code. An embodiment for determining the processing content of the stability processing based on the discontinuity determination code will be described.

(About encoding side)
FIG. 16 shows a functional configuration example of the encoder 2, and FIG. As shown in FIG. 16, the encoder 2 includes an LP analysis / encoding unit 21, a residual encoding unit 22, and a code multiplexing unit 23.

  Of these, FIG. 18 shows a functional configuration example of the LP analysis / encoding unit 21, and FIG. 19 shows a process flow diagram of the LP analysis / encoding unit 21. As shown in FIG. 18, the LP analysis / encoding unit 21 includes an LP analysis unit 210, an LP-ISF conversion unit 211, an ISF encoding unit 212, a discontinuity determination unit 213, an ISF concealment unit 214, and an ISF-LP conversion unit. 215 and an ISF buffer 216.

  In the LP analysis / encoding unit 21, the LP analysis unit 210 performs linear prediction analysis on the input signal to obtain a linear prediction coefficient (step T41 in FIG. 17 and step U41 in FIG. 18). In calculating the linear prediction coefficient, the Levinson-Durbin method or the like can be used after calculating the autocorrelation function from the speech signal.

  The LP-ISF conversion unit 211 converts the calculated linear prediction coefficient into an ISF parameter in the same manner as in the first embodiment (steps T42 and U42). A method described in non-patent literature may be used for conversion from the linear prediction coefficient to the ISF parameter.

  The ISF encoding unit 212 calculates the ISF code by encoding the ISF parameter by a predetermined method (steps T43 and U43), and converts the quantized ISF parameter obtained in the encoding process into the discontinuity determining unit 213, The data is output to the ISF concealment unit 214 and the ISF-LP conversion unit 215 (step U47). Here, the quantized ISF parameter is equal to the ISF parameter obtained by dequantizing the ISF code. As an encoding method, an error vector from an average vector determined by vector encoding, the ISF of the immediately preceding frame, or learning in advance may be encoded by vector quantization or the like.

と、量子化ISFパラメータ

を用いて、次式に従い不連続の判定を行い（ステップＴ４４、Ｕ４６）、その判定結果を不連続判定部２１３の内部バッファに格納する。ここで、Thres_ωは事前に定めた閾値、P’は次式を満たす整数である（0＜P’≦P）。

The discontinuity determination unit 213 encodes the discontinuity determination flag stored in an internal buffer (not shown) built in the discontinuity determination unit 213, and outputs the obtained discontinuity determination code (step U47). The discontinuity determination unit 213 also hides the hidden ISF parameter read from the ISF buffer 216.

And quantized ISF parameters

Is used to determine discontinuity according to the following equation (steps T44, U46), and the determination result is stored in the internal buffer of the discontinuity determination unit 213. Here, Thres _ω is a predetermined threshold value, and P ′ is an integer that satisfies the following equation (0 <P ′ ≦ P).

  Here, an example in which discontinuity determination is performed using the Euclidean distance between ISF parameters has been described, but discontinuity determination may be performed by another method.

  The ISF concealment unit 214 calculates concealment ISF parameters from the quantized ISF parameters by the same processing as the ISF concealment unit on the decoder side, and outputs the obtained concealment ISF parameters to the ISF buffer 216 (steps U44 and U45). . The processing procedure of the ISF concealment process may be any method as long as it is the same process as the packet loss concealment unit on the decoder side.

  The ISF-LP conversion unit 215 converts the quantized ISF parameter to calculate a quantized linear prediction coefficient, and outputs the obtained quantized linear prediction coefficient to the residual encoding unit 22 (step T45). As a method for converting the ISF parameter into the quantized linear prediction coefficient, a method described in non-patent literature may be used.

  The residual encoding unit 22 filters the speech signal using the quantized linear prediction coefficient and calculates a residual signal (step T46).

  Next, the residual encoding unit 22 encodes the residual signal by encoding means using CELP or TCX (Transform Coded Excitation), encoding means using switching between CELP and TCX, and the like. Output (step T47). Since the process of the residual encoding unit 22 is not related to the present invention, the description thereof is omitted.

  The code multiplexing unit 23 collects the ISF code, the discontinuity determination code, and the residual code in a predetermined order, and outputs the obtained speech code (step T48).

(Decryption side)
Since the speech signal processing apparatus according to the fourth embodiment has the same configuration as the speech decoder 1 of FIG. 1 described above and is characterized by the speech code decoding unit, the speech code decoding unit will be described below. FIG. 20 shows a functional configuration example of the speech code decoding unit 12D, and FIG. 21 shows a flow of LP coefficient calculation processing. The speech code decoding unit 12D of FIG. 20 has a discontinuity detection unit 129 added to the configuration of FIG. 3 described above.

  The ISF decoding unit 120 decodes the ISF code and outputs it to the stability processing unit 121 and the internal state buffer 14 (step S41 in FIG. 21).

  The discontinuity detection unit 129 decodes the discontinuity determination code, and outputs the obtained discontinuity detection result to the stability processing unit 121 (step S42 in FIG. 21).

  The stability processing unit 121 performs stability processing according to the discontinuity detection result (step S43 in FIG. 21). As the processing procedure of the stability processing unit, the same method as in the first and third embodiments can be used.

In addition to the discontinuity detection result obtained from the discontinuity determination code, the stability processing unit 121 may perform stability processing as follows based on other parameters included in the speech code. Good. For example, the stability processing unit 121 calculates the ISF stability stab according to the following equation, and if the ISF stability exceeds the threshold, the discontinuity detection result indicating that discontinuity is detected from the discontinuity determination code is obtained. Even if it is obtained, the stability processing may be performed as if discontinuity is not detected. Here, C is a predetermined constant.

  The ISF-ISP conversion unit 122A in the LP coefficient calculation unit 122 converts ISF parameters into ISP parameters by the same processing procedure as in the first embodiment (step S44 in FIG. 21).

  The ISP interpolation unit 122B calculates ISP parameters for each subframe by the same processing procedure as in the first embodiment (step S45 in FIG. 21).

  The ISP-LPC conversion unit 122C converts the ISP parameter calculated for each subframe into an LPC parameter by the same processing procedure as in the first embodiment (step S46 in FIG. 21).

  In the fourth embodiment as described above, the encoding side performs discontinuity determination (for example, discontinuous determination using the Euclidean distance between the concealed ISF parameter and the quantized ISF parameter), and auxiliary information regarding the determination result. Is output to the decoding side, and the decoding side estimates discontinuity using the auxiliary information obtained by decoding. In this way, appropriate processing according to the result of the discontinuity determination on the encoding side can be executed while the encoding side and the decoding side cooperate.

[Fifth Embodiment]
(About encoding side)
The functional configuration of the encoder is the same as the functional configuration of FIG. 16 according to the fourth embodiment, and the processing flow of the encoder is the same as the processing flow of FIG. 17 according to the fourth embodiment. Here, an LP analysis / encoding unit in a fifth embodiment different from the fourth embodiment will be described.

  FIG. 22 shows a functional configuration example of the LP analysis / encoding unit, and FIG. 23 shows a processing flow of the LP analysis / encoding unit. As illustrated in FIG. 22, the LP analysis / encoding unit 21S includes an LP analysis unit 210, an LP-ISF conversion unit 211, an ISF encoding unit 212, a discontinuity determination unit 213, an ISF concealment unit 214, and an ISF-LP conversion unit. 215 and an ISF buffer 216.

  In such an LP analysis / encoding unit 21S, the LP analysis unit 210 performs linear prediction analysis on the input signal by the same processing as in the fourth embodiment to obtain a linear prediction coefficient (step U51 in FIG. 23). ).

  The LP-ISF conversion unit 211 converts the calculated linear prediction coefficient into an ISF parameter by the same processing as in the fourth embodiment (step U52 in FIG. 23). A method described in non-patent literature may be used for conversion from the linear prediction coefficient to the ISF parameter.

  The ISF encoding unit 212 reads the discontinuity determination flag stored in the internal buffer (not shown) of the discontinuity determination unit 213 (step U53 in FIG. 23).

<When the discontinuity determination flag indicates discontinuity detection>
The ISF encoding unit 212 calculates an ISF code by vector quantization of the ISF residual parameter r _i calculated by the following equation (step U54 in FIG. 23). Here, the ISF parameter calculated by the LP-ISF converter is ω _i , and the average vector obtained in advance by learning is mean _i .

を用いてISF残差パラメータバッファを次式に従い更新する（図２３のステップＵ５５）。

Next, the ISF encoding unit 212 quantizes the ISF residual parameter r _i obtained by quantizing the ISF residual parameter r _i.

Is used to update the ISF residual parameter buffer according to the following equation (step U55 in FIG. 23).

とした。

<When the discontinuity determination flag does not indicate discontinuity detection>
The ISF encoding unit 212 calculates an ISF code by vector quantization of the ISF residual parameter r _i calculated by the following equation (step U54 in FIG. 23). Here, the ISF residual parameter obtained by decoding in the previous frame is

It was.

を用いて、ISF残差パラメータバッファを次式に従い更新する（図２３のステップＵ５５）。

Next, the ISF encoding unit 212 quantizes the ISF residual parameter r _i obtained by quantizing the ISF residual parameter r _i.

Is used to update the ISF residual parameter buffer according to the following equation (step U55 in FIG. 23).

  Through the above procedure, the ISF encoding unit 212 calculates the ISF code, and outputs the quantized ISF parameter obtained in the encoding process to the discontinuity determining unit 213, the ISF concealing unit 214, and the ISF-LP converting unit 215. To do.

  Similar to the fourth embodiment, the ISF concealment unit 214 calculates concealment ISF parameters from the quantized ISF parameters by the same processing as the ISF concealment unit on the decoder side, and outputs the concealment ISF parameters to the ISF buffer 216 (step U56 in FIG. 23). , U58). The processing procedure of the ISF concealment process may be any method as long as it is the same process as the packet loss concealment unit on the decoder side.

  The discontinuity determination unit 213 performs discontinuity determination by the same processing as in the fourth embodiment, and stores the determination result in an internal buffer (not shown) of the discontinuity determination unit 213 (step U57 in FIG. 23).

  The ISF-LP conversion unit 215 converts the quantized ISF parameter, calculates the quantized linear prediction coefficient, and outputs it to the residual encoding unit 22 (FIG. 16), as in the fourth embodiment ( Step U58 in FIG. 23).

(Decryption side)
Since the speech signal processing apparatus according to the fifth embodiment has the same configuration as the speech decoder 1 of FIG. 1 described above and has a feature in the speech code decoding unit, the speech code decoding unit will be described below. FIG. 24 shows a functional configuration example of the speech code decoding unit 12E, and FIG. 25 shows a flow of LP coefficient calculation processing. The speech code decoding unit 12E in FIG. 24 has a discontinuity detection unit 129 added to the configuration in FIG. 3 described above.

  The discontinuity detection unit 129 decodes the discontinuity determination code, and outputs the obtained discontinuity determination flag to the ISF decoding unit 120 (step S51 in FIG. 25).

  The ISF decoding unit 120 calculates ISF parameters as follows according to the value of the discontinuity determination flag, and outputs the ISF parameters to the stability processing unit 121 and the internal state buffer 14 (step S52 in FIG. 25).

、事前に学習で求めた平均ベクトルをmean_iとして、次式に従い量子化ISFパラメータ

を求める。

<When the discontinuity determination flag indicates discontinuity detection>
The ISF decoding unit 120 calculates a quantized ISF residual parameter obtained by decoding the ISF code.

Quantized ISF parameter according to the following equation, with mean _i obtained by learning as mean _i

Ask for.

Next, the ISF decoding unit 120 updates the ISF residual parameter stored in the internal state buffer 14 according to the following equation.

を内部状態バッファ１４から読み出し、
得られたISF残差パラメータ

、事前に学習で求めた平均ベクトルmean_i、及びISF符号を復号して得られる量子化ISF残差パラメータ

から、次式に従い量子化ISFパラメータ

を求める。

<When the discontinuity determination flag does not indicate discontinuity detection>
The ISF decoding unit 120 uses the ISF residual parameter obtained by decoding in the immediately preceding frame.

From the internal state buffer 14;
Resulting ISF residual parameters

, Average vector mean _i obtained by learning in advance, and quantized ISF residual parameter obtained by decoding ISF code

To quantized ISF parameters according to

Ask for.

Next, the ISF decoding unit 120 updates the ISF residual parameter stored in the internal state buffer 14 according to the following equation.

  The stability processing unit 121 performs the same processing as when the discontinuity described in the first embodiment is not detected (step S53 in FIG. 25).

  The ISF-ISP conversion unit 122A in the LP coefficient calculation unit 122 converts the ISF parameter into the ISP parameter by the same processing procedure as in the first embodiment (step S54 in FIG. 25).

  The ISP interpolation unit 122B calculates ISP parameters for each subframe by the same processing procedure as in the first embodiment (step S55 in FIG. 25).

  The ISP-LPC converter 122C converts the ISP parameter calculated for each subframe into an LPC parameter by the same processing procedure as in the first embodiment (step S56 in FIG. 25).

  In the fifth embodiment as described above, on the encoding side, when the discontinuity determination flag does not indicate discontinuous detection, the ISF residual parameter is obtained using the ISF residual parameter obtained by decoding in the immediately preceding frame. On the other hand, if the discontinuity determination flag indicates discontinuous detection, the use of the ISF residual parameter obtained by decoding in the immediately preceding frame is avoided. Similarly, on the decoding side, when the discontinuity determination flag does not represent discontinuous detection, the ISF residual parameter obtained by decoding in the immediately preceding frame is used to calculate the quantized ISF parameter, When the discontinuity determination flag indicates discontinuous detection, use of the ISF residual parameter obtained by decoding in the immediately preceding frame is avoided. In this way, appropriate processing according to the result of the discontinuity determination can be executed while the encoding side and the decoding side cooperate.

[Sixth Embodiment]
The first to fifth embodiments may be combined. For example, as described in the fourth embodiment, on the decoding side, when the discontinuity determination code included in the speech code from the encoding side is decoded to detect discontinuity, and the discontinuity is detected, the following is performed: You may carry out as follows.

については、第３実施形態に記載のとおり、低次P’次元（0＜P’≦P）までのISFパラメータを次式により置き換える。

ISF parameters stored in internal state buffer

As described in the third embodiment, the ISF parameters up to the low-order P ′ dimension (0 <P ′ ≦ P) are replaced by the following equations.

On the other hand, the ISF parameter of the current frame is calculated according to the following equation as described in the fifth embodiment.

  Thereafter, using the ISF parameters obtained as described above, LP coefficients are obtained by the processes of the ISF-ISP converter 122A, ISP interpolator 122B, and ISP-LPC converter 122C, as in the first embodiment.

  A mode in which the first to fifth embodiments are arbitrarily combined as described above is also effective.

[Seventh Embodiment]
In the first to sixth embodiments and the modifications described above, the situation of frame loss on the decoding side (for example, the situation of single frame loss or continuous frame loss) may be taken into account. In the seventh embodiment, the discontinuity detection may be performed by using the result of decoding the discontinuity determination code included in the speech code, for example, and the method is not limited to the above.

  Since the speech signal processing apparatus according to the seventh embodiment has the same configuration as the speech decoder 1 of FIG. 1 described above and is characterized by the speech code decoding unit, the speech code decoding unit will be described below.

  FIG. 26 shows a configuration example of the speech decoder 1S according to the seventh embodiment, and FIG. 27 shows a process flow diagram in the speech decoder. As shown in FIG. 26, the speech decoder 1S determines the packet reception state in the past several frames and stores the packet loss history in addition to the speech code decoding unit 12G, the concealment signal generation unit 13, and the internal state buffer 14 described above. A reception state determination unit 16 is provided.

  The reception state determination unit 16 determines the packet reception state, and updates the packet loss history information based on the determination result (step S50 in FIG. 27).

  When a packet loss is detected (NO in step S100), the reception state determination unit 16 outputs the packet loss detection result of the frame to the concealment signal generation unit 13, and the concealment signal generation unit 13 outputs the concealment signal described above. Are generated and the internal state is updated (steps S300 and S400). The concealment signal generation unit 13 may also use the packet loss history information for parameter interpolation or the like.

  On the other hand, if no packet loss is detected (YES in step S100), the reception state determination unit 16 obtains the packet loss history information including the packet loss detection result of the frame and the voice code included in the received packet. The speech code decoding unit 12 decodes the speech code and updates the internal state (steps S200 and S400).

  Thereafter, the processing of steps S50 to S400 is repeated until the end of communication (until YES is determined in step S500).

  FIG. 28 shows a functional configuration example of the speech code decoding unit 12G, and FIG. 29 shows a flow diagram of LP coefficient calculation processing. Here, an example in which the packet loss history information is used only for the LP coefficient calculation unit 122 will be described, but a configuration in which the packet loss history information is input to other components may be used.

  Since the configuration other than the configuration related to the LP coefficient calculation processing is the same as that of the first embodiment in the speech code decoding unit 12G, the configuration and operation related to the LP coefficient calculation processing will be described below.

  The ISF decoding unit 120 decodes the ISF code in the same manner as in the first embodiment, and outputs the ISF parameter to the stability processing unit 121 (step S71 in FIG. 29).

  The discontinuity detection unit 129 determines the reception state with reference to the packet loss history information (step S72). The discontinuity detection unit 129 stores in advance a specific reception pattern such as packet loss before 3 frames, normal reception before 2 frames, packet loss before 1 frame, and the corresponding reception pattern appears. In this case, the reception state flag may be turned off, and the reception state flag may be turned on in other cases.

  In addition, the discontinuity detection unit 129 performs discontinuity detection in the same manner as in any of the methods in the first to sixth embodiments described above.

  Then, the stability processing unit 121 performs stability processing according to the reception state flag and the result of discontinuity detection, for example, as follows (step S73).

  When the reception state flag is off, the stability processing unit 121 performs the same processing as when no discontinuity is detected regardless of the result of discontinuity detection.

  On the other hand, when the reception state flag is on and no discontinuity is detected as a result of the discontinuity detection, the stability processing unit 121 performs the same processing as when no discontinuity is detected.

  Furthermore, when the reception state flag is on and the result of discontinuity detection is that discontinuity is detected, the stability processing unit 121 performs the same processing as when discontinuity is detected.

  Thereafter, the operations (steps S74 to S76) of the ISF-ISP converter 122A, ISP interpolator 122B, and ISP-LPC converter 122C in the LP coefficient calculator 122 are the same as those in the first embodiment.

  In the seventh embodiment as described above, by performing stability processing according to the result of discontinuity detection and the state of the reception state flag, the situation of frame loss (for example, whether it is single frame loss or continuous frame loss) (Situation) can be taken into account and more accurate processing can be executed.

[Audio signal processing program]
Hereinafter, an audio signal processing program for causing a computer to operate as the audio signal processing apparatus according to the present invention will be described.

  FIG. 32 is a diagram showing various configuration examples of the audio signal processing program. FIG. 30 is a diagram illustrating a hardware configuration example of a computer, and FIG. 31 is an external view of the computer. Audio signal processing programs P1 to P4 (hereinafter collectively referred to as “audio signal processing program P”) shown in FIGS. 32A to 32D operate using the computer C10 shown in FIGS. 31 and 32 as an audio signal processing device. Can be made. Note that the audio signal processing program P described in this specification is not limited to the computer shown in FIGS. 31 and 32, and any information processing apparatus such as a mobile phone, a personal digital assistant, or a portable personal computer can be used. The audio signal processing program P can be operated.

  The audio signal processing program P can be provided by being stored in the recording medium M. The recording medium M is exemplified by a recording medium such as a flexible disk, CD-ROM, DVD, or ROM, or a semiconductor memory.

  As shown in FIG. 30, the computer C10 stores a reading device C12 such as a flexible disk drive device, a CD-ROM drive device, and a DVD drive device, a working memory (RAM) C14, and a program stored in the recording medium M. A memory C16 to be stored, a display C18, a mouse C20 and a keyboard C22 as input devices, a communication device C24 for transmitting and receiving data and the like, and a central processing unit (CPU) C26 for controlling execution of a program. .

  When the recording medium M is inserted into the reading device C12, the computer C10 can access the audio signal processing program P stored in the recording medium M from the reading device C12, and the audio signal processing program P serves as an audio signal processing device. It becomes possible to operate.

  As shown in FIG. 31, the audio signal processing program P may be provided via a network as a computer data signal W superimposed on a carrier wave. In this case, the computer C10 can store the audio signal processing program P received by the communication device C24 in the memory C16 and execute the audio signal processing program P.

  The audio signal processing program P can take various configurations shown in FIGS. These correspond to the configurations of claims 18 to 21 according to the audio signal processing program recited in the claims. For example, the audio signal processing program P1 illustrated in FIG. 32A includes a discontinuity estimation module P11 and a discontinuity correction module P12. The audio signal processing program P2 shown in FIG. 32B includes an ISF / LSF quantization module P21, an ISF / LSF concealment module P22, a discontinuity estimation module P23, and an auxiliary information encoding module P24. The audio signal processing program P3 shown in FIG. 32C includes a discontinuity estimation module P31, an auxiliary information encoding module P32, and an ISF / LSF quantization module P33. The audio signal processing program P4 shown in FIG. 32 (d) includes an auxiliary information decoding module P41, a discontinuity correction module P42, and an ISF / LSF decoding module P43.

  According to the various embodiments described above, it is possible to reduce discontinuous sounds that can occur when returning from a packet loss at the time of voice start and improve subjective quality.

  The stability processing unit, which is the first feature of the invention, increases the interval given between the elements of the ISF parameter, for example, when the discontinuity is detected in the packet normally received first after the packet loss. This can prevent the gain of the LP coefficient from becoming too large. Since it is possible to prevent both the gain of the LP coefficient and the power of the excitation signal from increasing, the discontinuity of the synthesized signal is reduced and the subjective quality deterioration is reduced. The stability processing unit may reduce the discontinuity of the composite signal by multiplying the composite signal by a gain calculated using an LP coefficient or the like.

  The discontinuity detector, which is the second feature of the invention, monitors the gain of the excitation signal included in the packet normally received first after the packet loss, and detects a packet in which the gain of the excitation signal increases more than a certain amount. Estimate discontinuities.

  The audio signal processing apparatus according to an embodiment of the present invention has a sudden increase in amplitude of decoded speech obtained as a result of speech packet decoding with respect to a speech packet normally received first after packet loss. A discontinuity estimator that estimates the generated discontinuity and a discontinuity corrector that corrects the discontinuity of the decoded speech are provided.

  The discontinuity estimator may estimate the discontinuity of the decoded speech based on the power of the excitation signal.

  Further, the discontinuity estimator may estimate the discontinuity of the decoded speech based on a quantized codebook gain used for excitation signal calculation.

  The speech signal processing apparatus further includes an auxiliary information decoder that decodes auxiliary information related to discontinuity transmitted from the encoder, and the discontinuous estimator includes the auxiliary information decoder that receives the auxiliary information code. You may estimate the discontinuity of decoded audio | voice using the auxiliary information regarding the discontinuity output after decoding.

  The discontinuity corrector may correct an ISF parameter or an LSF parameter (hereinafter referred to as “ISF / LSF parameter”) in accordance with the discontinuity estimation result.

  More specifically, the discontinuity corrector described above may change the interval between each element of the ISF / LSF parameter given to guarantee the stability of the synthesis filter in accordance with the discontinuity estimation result. .

  At this time, the above discontinuity corrector may provide an interval between each element of the ISF / LSF parameter given to guarantee the stability of the synthesis filter, which is larger than the interval usually given to guarantee the stability. Good.

  In addition, the discontinuity corrector described above gives the interval obtained by equally dividing the ISF / LSF parameter up to a predetermined dimension, and the interval between each element of the ISF / LSF parameter to guarantee the stability of the synthesis filter. It may be used as

  Further, the discontinuity corrector described above may replace some or all of the ISF / LSF parameters with a predetermined vector.

  An audio signal processing apparatus according to an embodiment of the present invention includes an ISF / LSF quantizer that quantizes an ISF / LSF parameter, and an ISF / LSF that generates concealed ISF / LSF parameters that are concealment information related to the ISF / LSF parameters. Packet loss occurs using the distance between the concealer and the quantized ISF / LSF parameter obtained in the quantization process of the ISF / LSF quantizer and the concealed ISF / LSF parameter generated by the ISF / LSF concealer. And a discontinuity estimator that estimates discontinuity that occurs first in a voice packet that is normally received first, and an auxiliary information encoder that encodes auxiliary information related to the discontinuity.

  An audio signal processing device according to an embodiment of the present invention encodes a discontinuity estimator that estimates discontinuity that occurs in a voice packet that is normally received first after packet loss and auxiliary information related to the discontinuity. If the discontinuity estimation is not performed by the auxiliary information encoder to be converted and the discontinuity estimator, the past quantized ISF / LSF residual parameter is used for ISF / LSF quantization in the frame, and the discontinuity estimation is performed. An ISF / LSF quantizer that avoids using past quantized ISF / LSF residual parameters for ISF / LSF quantization in the frame when the unit estimates discontinuity.

  An audio signal processing apparatus according to an embodiment of the present invention includes: an auxiliary information decoder that decodes and outputs auxiliary information related to discontinuity that occurs in an audio packet that is normally received first after packet loss; If the discontinuity corrector that corrects the discontinuity and the discontinuity is not estimated by the output of the auxiliary information decoder, the past quantized ISF / LSF residual parameter is used for ISF / LSF calculation in the frame, An ISF / LSF decoder that avoids using past quantized ISF / LSF residual parameters for ISF / LSF calculation in the frame when discontinuity is estimated by the output of the auxiliary information decoder .

  The audio signal processing apparatus further includes a reception state determination unit that determines the packet reception state of a predetermined number of past frames, and the discontinuity corrector also includes a packet reception state determination result in addition to the discontinuity estimation result. As a basis, a configuration for correcting discontinuity may be adopted.

  Now, the audio signal processing apparatus according to an embodiment of the present invention can be regarded as an invention related to an audio signal processing method and an invention related to an audio signal processing program, and can be described as follows.

  An audio signal processing method according to an embodiment of the present invention is an audio signal processing method executed by an audio signal processing device, and an audio packet is first received normally after a packet loss has occurred. A step of estimating a discontinuity caused by a sudden increase in amplitude of the decoded speech obtained as a result of decoding, and a step of correcting the discontinuity of the decoded speech.

  An audio signal processing method according to an embodiment of the present invention is an audio signal processing method executed by an audio signal processing device, which is a step of quantizing ISF / LSF parameters and concealment information related to ISF / LSF parameters. Packet loss occurred using the step of generating the concealed ISF / LSF parameter and the distance between the quantized ISF / LSF parameter obtained in the quantization process of the ISF / LSF parameter and the generated concealed ISF / LSF parameter. A step of estimating a discontinuity that occurs in a voice packet that is normally received first and a step of encoding auxiliary information related to the discontinuity.

  An audio signal processing method according to an embodiment of the present invention is an audio signal processing method that is executed by an audio signal processing device, and includes discontinuities that occur in an audio packet that is normally received first after a packet loss has occurred. If the discontinuity is not estimated, the past quantized ISF / LSF residual parameters are used for ISF / LSF quantization in the frame, and A step of avoiding using past quantized ISF / LSF residual parameters for ISF / LSF quantization in the frame if continuity is estimated.

  An audio signal processing method according to an embodiment of the present invention is an audio signal processing method executed by an audio signal processing apparatus, and relates to discontinuity that occurs in an audio packet that is normally received first after a packet loss has occurred. A step of decoding and outputting auxiliary information; a step of correcting discontinuity of decoded speech; and if the auxiliary information does not indicate an estimation of discontinuity, a past quantized ISF / LSF residual parameter When the auxiliary information indicates the estimation of discontinuity in the ISF / LSF calculation in the step, avoiding using the past quantized ISF / LSF residual parameter in the ISF / LSF calculation in the frame, Prepare.

  An audio signal processing program according to an embodiment of the present invention causes a computer to rapidly increase the amplitude of decoded audio obtained as a result of audio packet decoding with respect to an audio packet that is normally received first after packet loss. This is a speech signal processing program for operating as a discontinuity estimator that estimates discontinuities caused by the above and a discontinuity corrector that corrects discontinuities in decoded speech.

  An audio signal processing program according to an embodiment of the present invention generates a concealed ISF / LSF parameter that is concealment information related to an ISF / LSF parameter, and an ISF / LSF quantizer that quantizes the ISF / LSF parameter. Using the distance between the ISF / LSF concealer, the quantized ISF / LSF parameter obtained in the quantization process of the ISF / LSF quantizer, and the concealed ISF / LSF parameter generated by the ISF / LSF concealer, the packet Voice signal to operate as a discontinuity estimator that estimates discontinuity that occurs in the first normally received voice packet after loss and an auxiliary information encoder that encodes auxiliary information related to the discontinuity It is a processing program.

  An audio signal processing program according to an embodiment of the present invention includes a discontinuity estimator that estimates a discontinuity that occurs in a voice packet that is normally received first after a packet loss, and a discontinuity assistance. When the auxiliary information encoder that encodes information and the discontinuity estimator does not estimate discontinuity, the past quantized ISF / LSF residual parameter is used for ISF / LSF quantization in the frame, and When the discontinuity estimator estimates discontinuity, to operate as an ISF / LSF quantizer that avoids using past quantized ISF / LSF residual parameters for ISF / LSF quantization in the frame This is an audio signal processing program.

  An audio signal processing program according to an embodiment of the present invention includes an auxiliary information decoder that decodes and outputs auxiliary information related to discontinuity that occurs in an audio packet that is normally received first after a packet loss, If the discontinuity corrector that corrects the discontinuity of the decoded speech and the discontinuity is not estimated by the output of the auxiliary information decoder, the ISF / LSF calculation of the past quantized ISF / LSF residual parameter in the current frame In the case where the discontinuity is estimated by the output of the auxiliary information decoder, an ISF / LSF decoder that avoids using past quantized ISF / LSF residual parameters for ISF / LSF calculation in the frame, Is an audio signal processing program for operating as above.

DESCRIPTION OF SYMBOLS 1, 1S, 1X ... Voice decoder, 11 ... Packet loss detection part, 12, 12A, 12B, 12D, 12E, 12G, 12S ... Voice code decoding part, 13, 13X ... Concealment signal generation part, 14 ... Internal state buffer 15... Normalized prediction residual power decoding unit, 16... Reception state determination unit, 21, 21 S... Analysis / encoding unit, 22... Residual encoding unit, 23. 121, 121S, stability processing unit, 121X, gain adjustment unit, 121Y, gain multiplication unit, 122, LP coefficient calculation unit, 122A, ISF-ISP conversion unit, 122B, ISP interpolation unit, 122C, ISP-LPC conversion unit, 123: Adaptive codebook calculation unit, 124 ... Fixed codebook decoding unit, 125 ... Gain decoding unit, 126 ... Excitation vector synthesis unit, 127 ... Post filter, 128 ... Synthesis filter, 129 ... Discontinuity detection unit, 130 ... LP coefficient interpolation unit 131... Pitch lag interpolation unit 132... Gain interpolation unit 133 133 noise signal generation unit 134... Post filter 135 135 synthesis filter 136 136 adaptive codebook calculation unit 137 excitation vector synthesis unit 138 ... Normalized prediction residual adjustment unit, 210 ... LP analysis unit, 211 ... LP-ISF conversion unit, 212 ... ISF encoding unit, 213 ... Discontinuity determination unit, 214 ... ISF concealment unit, 215 ... ISF-LP conversion unit 216 ... ISF buffer, C10 ... computer, C12 ... reading device, C14 ... working memory, C16 ... memory, C18 ... display, C20 ... mouse, C22 ... keyboard, C24 ... communication device, C26 ... CPU, M ... recording medium , P1 to P4 ... audio signal processing program, P11 ... discontinuity estimation module, P12 ... discontinuity correction module, P21 ... ISF / LSF quantization module P22 ... ISF / LSF concealment module, P23 ... Discontinuous estimation module, P24 ... Auxiliary information encoding module, P31 ... Discontinuous estimation module, P32 ... Auxiliary information encoding module, P33 ... ISF / LSF quantization module, P41 ... Assist Information decoding module, P42 ... discontinuity correction module, P43 ... ISF / LSF decoding module, W ... computer data signal.

Claims (3)

An audio signal processing method executed by an audio signal processing device,
A decoding step of decoding auxiliary information related to discontinuity of decoded speech obtained as a result of speech packet decoding, which is auxiliary information transmitted from the encoder;
Using the auxiliary information decoded by the decoding step, estimating the discontinuity of the decoded speech; and
A correction step that corrects the discontinuity of the decoded speech when it is estimated that the estimation step is discontinuous, and a correction step that does not correct the discontinuity of the decoded speech when it is not estimated that the estimation step is discontinuous When,
Including
The discontinuity is caused by a sudden increase in the amplitude of decoded speech obtained as a result of speech packet decoding with respect to the speech packet normally received first after packet loss.
Audio signal processing method. And further comprising a detecting step of detecting a packet loss of the transmitted voice packet,
The correction step corrects the discontinuity of the decoded speech when a packet loss is detected by the detection step and when the estimation step estimates that the packet loss is discontinuous,
The audio signal processing method according to claim 1. An auxiliary information decoder that decodes auxiliary information that is auxiliary information transmitted from the encoder and that is related to discontinuity in decoded speech obtained as a result of speech packet decoding;
A discontinuity estimator that estimates discontinuity of the decoded speech using the auxiliary information decoded by the auxiliary information decoder;
The discontinuity of the decoded speech is corrected when the discontinuity estimator is estimated to be discontinuous, and the discontinuity of the decoded speech is corrected when the discontinuity estimator does not estimate the discontinuity. A non-correcting discontinuity corrector;
With
The discontinuity is caused by a sudden increase in the amplitude of decoded speech obtained as a result of speech packet decoding with respect to the speech packet normally received first after packet loss.
Audio signal processing device.

Images