KR20040001733A - Apparatus for calculating an Observation Probability for a search of hidden markov model - Google Patents

Abstract

PURPOSE: An apparatus for operating observation probability for searching a hidden markov model is provided to efficiently execute observation probability operation executing the most operations, and reduce power consumption. CONSTITUTION: A storage device stores the mean of parameters extracted from representative phonemes and a distribution degree of the mean. A subtractor(405) obtains a difference between the mean and a feature extracted from a voice signal to be a recognition object. A multiplier(406) multiplies output of the subtractor by the distribution degree provided from the storage device. A squarer(407) squares the result of the multiplication of the multiplier. An accumulator(408) accumulates output of the squarer.

Description

Apparatus for calculating an Observation Probability for a search of hidden markov model}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to speech recognition devices and, more particularly, to an observation probability computing device for searching hidden Markov models.

The voice recognition function can be applied to almost all electronic products that humans can encounter in everyday life, and the application trend is expected to be gradually extended to low-cost e-toy products.

The first company to offer usage technology for speech recognition proved its effectiveness by first applying hidden Markov models to character recognition. (1997.06, US 5,636,291)

These speech recognition methods are largely divided into three parts: pre-processor, front-end, and modeling. The pre-processor part is a step of recognizing the lexicon of the character to be processed.

The front-end part has a function of extracting feature values or parameters to be compared from the recognized lexicons.

On the other hand, in the modeling part, based on the extracted parameters, a model that becomes an accurate criterion for the recognized character in the future is constructed through a training phase. In addition, based on the recognized vocabulary, a function of determining which of the predetermined characters is determined as the recognized character is performed.

Since then, IBM has disclosed speech recognition systems and methods using hidden Markov models that can be used in a wider range of applications. (1998.08, US 5,799,278) This technique uses hidden Markov models in the speech recognition process for isolated words, trained to recognize other words phonetically, and uses hidden Markov models that are suitable for recognizing many words. Method and voice recognition system.

In implementing such a speech recognition apparatus, it is required to shorten the computation time required for speech recognition. It has been observed that in the speech recognition apparatus using the hidden mark model, the computation for the observation probability calculation uses about 62%. Therefore, it is required to improve the computation speed.

An object of the present invention is to provide an improved observation probability calculating device for meeting the above requirements.

1 is a block diagram showing the configuration of a general speech recognition system.

Fig. 2 is a schematic view showing a method of obtaining a state string for any syllable.

3 is a schematic view of a process for word recognition.

4 is a block diagram showing the configuration of the apparatus for calculating the probability of observation according to the present invention.

5 is shown to help understand the selection of the bit resolution.

6 is shown to show an application example of the observation probability calculation device according to the present invention.

FIG. 7 is a block diagram schematically illustrating a process of receiving a control command and data in the apparatus shown in FIG. 6.

FIG. 8 is a timing diagram schematically showing a process of receiving a control command and data in the apparatus shown in FIG. 6.

Observation probability calculation apparatus according to the present invention to achieve the above object

A storage device for storing a mean of a parameter extracted from representative phonemes and a degree of distribution of the mean value (1 / σ);

A subtractor for calculating a difference between an average provided from the storage device and a feature extracted from a speech signal to be recognized;

A multiplier for multiplying the output of the subtractor with the degree of distribution provided from the storage device;

A multiplier for square operation of the multiplication result of the multiplier; And

And an accumulator accumulating the output of the squarer.

In the observation probability computing device of the present invention, it is preferable to further include registers for buffering the variance, the average, and the parameters, respectively.

In the observation probability calculating device of the present invention, it is more preferable to further include a register for buffering the accumulating result of the accumulator. Hereinafter, the configuration and operation of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram showing the configuration of a general speech recognition system. In the apparatus shown in FIG. 1, the A / D block 101 converts a voice signal input as a continuous signal into a digital signal that is easy to operate.

The pre-emphasis block 102 functions to emphasize high frequency components in order to clarify the distinction of pronunciation due to the nature of speech. The audio signal converted into a digital signal is separated and processed in a unit of a predetermined sampling number, which is divided into 240 samples (30 ms).

As the feature vector used in the current hidden Markov model is generally used the cepstrum and the energy generated from the spectrum, a calculation is required. Energy Calculation block 103. Here, to find the energy, we calculate the instantaneous energy for 30 영역 using the energy calculation formula in the timing domain. This calculation is the same as Equation 1.

<Equation 1>

This energy value is used to determine whether the currently input signal is a voice signal or a noise. In addition, fast Fourier transform, which is frequently used in signal processing, is used to obtain spectrum in the frequency domain. This spectrum is obtained through a 256 point FFT operation. This FFT operation performs a 256 Point Complex-FFT operation and may be represented by Equation 2.

<Equation 2>

After the energy calculation result is used to determine whether it is a voice signal or noise, if it is found to be a voice signal, the start and end of the voice should be determined. The function of determining the start and end of the voice signal is performed in the FineEndPoint block 104. When a valid word is determined as such, only the corresponding spectrum data is stored in the buffer. Therefore, the buffer block 105 stores only a valid voice signal in which the noise part of the speech word of the speaker is excluded.

In the Mel-filter block 106, a Mel-filtering operation for filtering with 32 bandwidths is performed as a preprocessing process for obtaining the Cepstrum from the spectral values.

This process yields spectral values for the 32 bands. By converting this value in the frequency domain back to the time domain, we can obtain the Cepstrum, a parameter used by the hidden Markov model. Inverse Discrete Cosine Transform (IDCT) operation is performed to convert to this time domain. (IDCT Block, 107)

These cepstrum and energy values are used for retrieval using the hidden Markov model, because the difference between the energy value and the cepstrum is too large (difference of the power of 10). Do. The scaling of this value is performed in Scale block 108. Adjust the value using Logarithm. In addition, the Cepstral Window block 109 separates the periodicity and the energy from the Mel-cepstrum value and works to improve the noise characteristics. Here, the improvement of the noise characteristic is calculated using Equation 3.

<Equation 3>

Here, Sin_Table may be configured as shown in Equation 4.

<Equation 4>

When the above operation is completed, the normalization block 110 as shown below normalizes the energy values, which are the ninth data of each frame, to a value within a predetermined range.

<Equation 5>

As shown in Equation 5, the largest value is found among the ninth data of each frame, and this value is subtracted from the energy data of all frames as shown in Equation 6 to obtain Normalized Energy.

<Equation 6>

In general, a method of increasing the type of a parameter (feature value) to increase the recognition rate is frequently used. The most commonly used method is to take the difference of the feature value between the frame and the frame as another feature value in addition to the feature value of each frame. The dynamic feature block 111 calculates the Delta Cepstrum and selects it as the secondary feature value. A method of calculating the difference between these cepstrums is shown in Equation (7).

<Equation 7>

In general, the frame to be computed is two frames each. When this operation is completed, the same number of Delta Cepstrums are generated as Cepstrum.

Through the above operation, feature values that are the target of hidden Markov model search are extracted.

From these feature values, a word search operation is performed using a predetermined hidden Markov model. The exploration of hidden Markov models involves three steps. First is the Observation Probability calculation block 112. Basically, the search and decision process is based on probability. It is probable to find the closest syllable. Probability values are largely divided into Observation Probability and Transition Probability. By accumulating these probability values, a sequence of syllables having the largest probability value is selected. The observation probability may be expressed as shown in Equation 8.

<Equation 8>

Here, dbx is a stochastic distance between the reference mean value and the feature value extracted from the input signal. The closer the probability is, the greater the probability is. The formula for calculating such a stochastic distance is shown in Equation 9.

<Equation 9>

Here, m represents an average value of the parameter, and Feature means a parameter value extracted from the input signal. p is the precision value and represents the distribution degree (variance 1 / σ²), and lw is the log weight. i represents a Mixture representing a representative type of phoneme. For example, in order to improve the accuracy of recognition, it is necessary to obtain representative values from many people, and classify these representative values into several groups representing a common type for a phoneme. Becomes k represents the number of frames and j represents the number of parameters. For reference, the number of frames varies according to the type of word, and Mixture can be classified into various types according to the pronunciation type of a general person. Log Weight is subtracted as the weight calculation in the linear domain is converted into the weight calculation in the log domain.

The observed probability calculated as described above is a probability that a phoneme of each pre-selected syllable can be observed, and each phoneme has a different probability value. Therefore, when the observation probability of all the phonemes is determined, the most suitable phoneme sequence is obtained by applying the same to the predetermined state machine sequence block 113. In general, each state machine of a hidden Markov model for independent word recognition is a sequence based on feature values for each phoneme of a word to be recognized.

FIG. 2 is a diagram schematically illustrating a method of obtaining a state string for an arbitrary syllable, and illustrates a state machine sequence for an arbitrary syllable “k”.

Assuming that the syllable "k" is composed of three sequential state strings (S1, S2, S3), the process of starting from the initial state S0 in FIG. 2 and finally reaching S3 via S1 and S2 It is shown. In FIG. 2, going to the right in the same state means a delay state, and the delay state is speaker dependent. That is, in some cases the syllable “k” may occur very short in time, but in other cases it may occur in a relatively long time. The longer a syllable occurs, the longer the delay in each state. In FIG. 2, Sil represents silent sound.

If the user pronounces "k", this status string will have the highest probability value. Therefore, many state strings as shown in FIG. 2 exist and a large amount of computation is required because a probability operation is performed on one input signal for each state string.

Finally, when a probabilistic operation (processing state-by-phoneme state processing for each phoneme) is completed, probability values are stored in a state machine of the final phoneme. In FIG. 2, a criterion for progressing each state stage is to obtain an Alpha value while selecting a branch that is maximum using Equation 10 below. These alpha values are the accumulated values of observation probabilities, and are obtained by using the transition probability between phonemes obtained from previous observation probability values and previous empirical experiments.

<Equation 10>

Here, State.Alpha is a probability value newly calculated and accumulated, and State.Alpha_prev is a probability value accumulated before. In addition, trans_prob [0] is a probability of transition from state Sn to Sn (for example, S0-> S0) and trans_prob [1] is a probability of transition from state Sn to state Sn + 1. O_prob is the observed probability calculated in the current state. In the Find Maximum Likelihood block 114, a function of selecting a recognized word based on a final cumulative probability value for each phoneme is performed as shown in Equation 10. In this case, the word having the largest probability value is selected as the recognized word.

The procedure for recognizing the word "KBS" will be described by way of example.

The word "KBS" consists of three syllables "K", "B", and "S". Also, the syllable "K" consists of three phonemes: "K", "E", and "Y". The syllable "B" consists of the phonemes "B" and "Y", and the syllable "S" consists of three phonemes "Y", "E", and "S".

The word "KBS" consists of eight phonemes: "k", "e", "yi", "bro", "yi", "yi", "e", and "su". The probability of observation and the transition between each phoneme are recognized.

In other words, in order to recognize the word "KBS", the eight phonemes "K", "E", "Y", "B", "Y", "Y", "E", and "S" are as accurate as possible. It should be recognized, and on the basis of it, the word "KBS" should be selected, as closely as possible the sequence between each phoneme.

The observation probability of each phoneme is calculated for the input voice signal. In order to calculate the observation probability, the similarity with the representative phonemes stored in the database, that is, the probability is calculated, and the probability for the representative phone with the highest probability becomes the observation probability. For example, for the phone "k", all of the representative phonemes stored in the database are compared and the representative phone "k" with the highest probability is selected.

When the observation probability is calculated for each phoneme of the input voice signal, that is, the representative phonemes for each phoneme of the voice signal are determined, the most suitable sequence is applied by applying the input voice signal to a state machine sequence composed of these phonemes. Will be determined. The state machine sequence is made up of eight phonemes: "K", "E", "Y", "B", "Y", "Y", "E", and "S", and each phoneme is observed. The word "KBS" having the highest probability and the cumulative value thereof is selected. Each phoneme consists of three states in detail.

3 is a schematic view of a process for word recognition.

For example, in order to recognize the word "KBS", each phoneme "k", "e", "yi", "bro", "yi", "yi", through the observation probability calculation block 112 is used. Observation probabilities for eight phonemes "E" and "S" are calculated, and the word "KBS", which is the word having the largest observation probability and the cumulative value of each phoneme, is selected through the state machine 113. will be.

In general, many existing speech recognition products design the above functions in software (C / C ++ language) or machine code (Assembly Code) to perform functions using a general purpose processor.

Another form of use is implemented by dedicated hardware (ASIC). Each of these two methods has advantages and disadvantages. Software processing takes a relatively long time, but its flexibility makes it easy to change functionality.

On the other hand, the method of processing with dedicated hardware has a relatively high processing speed and low power consumption compared to processing with software, but it is inflexible and thus cannot be changed.

Therefore, the present invention proposes a device that can support a relatively fast processing speed while being suitable for a software method whose function can be easily changed.

Table 1 shows the number of operations required to perform each function when using a general-purpose processor as a software processing method. The number of operations here is not the number of instructions but the number of operations such as multiplication, addition, logarithm, and exponentiation.

TABLE 1

calculate Pre-processing Mel-filtering & Cepstrum HMM Sum Pre-emphasis EnergyCalc. FFT Mel-Filter IDCT Scaling Cepstr. Observ.Prob. Statemachine multiplication 160 240 4096 234 288 9 36 43200 0 48,263 addition 160 239 6144 202 279 0 One 45600 600 53,225 Division 0 One 0 0 0 0 9 0 0 10 Square root 0 One 0 0 0 0 0 0 0 One LOG 0 0 0 32 0 0 0 0 One 33 Total operation 320 481 10240 468 567 9 46 88800 601 101,532

As shown in Table 1, the total number of operations required for general speech recognition processing is about 100,000, of which about 88.8% are observation probability operations.

In fact, when the above-described algorithm is implemented and executed on the ARM Processor series, which are widely used today, the total number of instructions for performing the entire function is about 36 million instructions, of which about 33 million instructions are required for searching hidden Markov models. It was analyzed. Table 2 categorizes the instructions required to perform voice recognition using the actual ARM processor by function block.

TABLE 2

Function block Number of instruction cycles Percentage (%) Observation Probability Operation 22,267,200 61.7% State machine update 11,183,240 30.7% FFT 910,935 2.50% Find Maximum Likelihood 531,640 1.46% Mel-Filtering / IDCT / Scaling 473,630 1.30% Dynamic Feature 283,181 0.78% Pre-emphasis & Energy Calculation 272,037 0.75% Cepstral Window & Normalize 156,061 0.43% Find End Point 123,050 0.30% Total 36,400,974 100.00%

As can be seen from Table 2, it takes about 62% to execute the command. Therefore, by supporting the observation probability operation portion that executes the most instructions as a dedicated device, it is possible to improve processing speed and reduce power consumption.

Accordingly, the present invention proposes a dedicated device capable of performing such an Observation Probability operation with fewer instructions, that is, fewer cycles.

In order to improve the calculation ability of the observation probability, the present invention provides a dedicated device capable of performing operations such as Equation 10 to Equation 10, which are the most computationally probabilistic distance calculations.

<Equation 11>

Where p [i] [j] represents the degree of distribution (dispersion, 1 / σ²) in precision, mean [i] [j] is the mean value of each phoneme, and feature [k] [j] is the parameter for the phoneme Values represent energy and cepstrum. In Equation 11, mean [i] [j]-feature [k] [j] indicate how far a parameter (distance) from a stochastic input phoneme differs from a predefined representative parameter. Take the square to calculate And multiply this by the variance to predict the objective real distance. Here, the representative parameter values are empirically obtained through numerous voice data, and the more voice data obtained from various people, the better the recognition rate.

However, in the present invention, in order to increase the recognition rate as much as possible in consideration of the limited characteristics of the hardware, that is, the limitation of the data bits (16 bits), an operation as shown in Equation 12 is performed.

<Equation 12>

Here, p [i] [j] is 1 / σ indicating the degree of distribution, unlike variance 1 / σ 2 in equation VII. The reason why the distribution degree 1 / σ is used instead of the variance 1 / σ 2 is as follows.

According to Equation 9, the result of squared operation of (m [i] [j] -feature [i] [j] and p [i] [j] is multiplied, but according to Equation 12, p [i] [j ] The result of calculating (m [i] [j] -feature [i] [j]) is squared.

Equation (9) requires the same bit resolution as the result of the squared operation to express p [i] [j], but according to Equation (12) (m [i] [j] -feature [i] [j] This means that only a bit resolution of the same level is required.

In other words, in order to maintain 16-bit bit resolution, 32 bits are required to express p [i] [j] according to Equation 9, but only 16 bits are used to express p [i] [j] according to Equation 12. This is necessary. On the other hand, according to Equation 12, the result of calculating p [i] [j] · (m [i] [j] -feature [i] [j]) is squared. Similar effects can be achieved with / σ².

4 is a block diagram showing the configuration of the apparatus for calculating the probability of observation according to the present invention. The apparatus shown in FIG. 4 includes a subtractor 405, a multiplier 406, a square 407, and an accumulator 408. Reference numerals 402, 403, 404, and 409 denote registers.

The external storage device 401 is a databased storage device that stores presion, mean, and features for all representative phonemes. Here, precision indicates distribution degree (1 / σ), mean is the average value of parameters (energy and cepstrum) representing each representative phoneme, and feature [k] [j] is the parameter value for phoneme. Means 켑 strum.

In the apparatus shown in FIG. 3, the difference between the mean and the feature is first obtained by using the subtractor 405, and the result is multiplied by the dispersion degree (1 / σ) through the multiplier 406 to obtain the actual distance. The result is to find the square through the square 407 to find the absolute difference, and to use the adder 408 to accumulate with the previous parameter.

That is, the result to be expressed in equation (12) is obtained by the multiplier (406), and is expressed in equation (9). The result of the operation is obtained at the accumulator 408.

P [i] [j], mean [i] [j], and feature [i] [j] are stored in the external storage device, and they are sequentially stored in the registers 402, 403, and 404 in a predetermined order. Is provided. The predetermined order is set such that i and j are sequentially increased.

p [i] [j], mean [i] [j], and feature [i] [j] are sequentially provided to the registers 402, 403, 404, changing i, j, and register 409. The final cumulative observation probability is obtained from.

By the cumulative calculation of these probabilities, the most probabilistic phonemes have the highest value. The registers 402, 403, 404, 409 at the beginning and the end of the operation are used for stabilization of the data.

In the apparatus illustrated in FIG. 4, the bit resolution of data may vary depending on the structure of the processor, and as the number of bits increases, detailed calculation results may be obtained. However, since the bit resolution is related to the size of the circuit, an appropriate resolution should be selected in consideration of the recognition rate.

5 is shown to help understand the selection of the bit resolution. As an example of the selection of the bit resolution, Figure 5 shows the internal bit resolution for a processor with 16 bit resolution. The truncation process for each step is based on the 16-bit data width limit and is the choice to prevent performance degradation as much as possible. By using the device proposed in the present invention, much improvement in processing speed can be achieved compared to the case of using only a general-purpose processor.

The feature and mean each consist of a 4-bit integer and a 12-bit prime. These features and mean are subtracted through the subtractor 405 to obtain a result value that is also composed of an integer of 4 bits and a decimal number of 12 bits.

precision consists of an integer of 7 bits and a decimal number of 9 bits. The result of the precision and the subtraction of the subtractor 405 is multiplied by the multiplier 406 to obtain a result value consisting of an integer of 10 bits and a decimal number of 6 bits.

The result of multiplying the multiplier 406 by the power of the multiplier 407 to obtain a result of 20-bit integer and 12-bit decimal, and add and scale the result of the 21-bit You get a result consisting of an integer and a decimal number of 11 bits.

Table 3 compares and analyzes the case in which the speech recognition algorithm using the commonly used hidden Markov model is executed in the general purpose processor (ARM Series) and the dedicated processor adopting the apparatus for calculating the probability of observation presented in the present invention. will be.

TABLE 3

Processor Cycle number Time] (20M CLK) ARM Processor 36,400,974 1.82 s Observation Probability Calculator 15,151,534 0.758 s

In Table 3, a general-purpose processor takes about 36 million cycles to perform voice recognition, while a dedicated processor employing a dedicated device can perform the required function in about half a million cycles. This allows for near real-time speech recognition, which means that even at a lower clock frequency, the same performance as a general-purpose processor can have significant effects on power consumption. For reference, the relationship between the power consumption and the clock frequency may be expressed as in Equation 13.

<Equation 13>

Here, P is a power consumption amount, C represents the capacitance value constituting the circuit. f represents the total degree of transition of the signal in the circuit, which is almost always the clock speed. V is the supply voltage. Therefore, if you lower the clock speed in half, you can theoretically cut power consumption in half.

As shown in FIG. 4, the apparatus stores the average parameters and transition probability values of representative phonemes for each type of person obtained by an empirical method in advance, and the degree of distribution and parameters extracted from the newly inputted voice. Such data is stored in the registers 302, 303, and 304 inside the dedicated device, which is related to power consumption in order to minimize signal changes caused by external data changes. The parameter extracted from the input voice among the data stored in the internal register and the average parameter Mean stored in advance are subjected to a subtraction operation through the subtractor 405 to obtain a difference therebetween.

This result is multiplied by Precision, which represents the degree of variance (1 / σ), by multiplier 406 and again by the square 407 to calculate the actual stochastic distance. Since this value is calculated only the current parameter in time among the many speech parameter frames forming the word, it should be accumulated by adding up with the stochastic distance previously calculated by the adder 408. A register 409 is used in addition to the adder 408 for the cumulative operation and the data stored in the register is provided to the adder 408 for the next operation.

These registers should be used not only for cumulative operation but also for minimizing signal transitions. The above process is applied equally to each predetermined phoneme, and the value is stored in the corresponding storage location for each phoneme / state. As a result, when the calculation for all the parameters of the input word is completed, the largest value among the cumulative values for each phoneme of each word may be recognized as the probability most similar word. Determining the final recognized word using the accumulated values will be performed by the existing processor.

6 is shown to show an application example of the observation probability calculation device according to the present invention. The apparatus shown in FIG. 6 uses a 3-bus system scheme as a speaker-independent speech recognition dedicated processor. The observation probability computing device according to the present invention is implemented inside the HMM module 628 shown in FIG. 4, and each component module is used for three buses for data (two read buses and one write bus). Share two OPcode buses.

In FIG. 6, a control unit (Ctrl Unit) 602 means a general-purpose processor, a REG FILE (register file) 604 means a module that performs a register file function, and the ALU 506 is an Arithmatic Logic. The MAC 608 means a multiply and ACcummulate function, the B SHIFTER (barrel shifter, 610) means a module performing a Barrel SHIFT function, and the FFT 612 is an FFT. SQRT 514 denotes a module performing a square and root arithmetic function, TIMER 616 denotes a module performing a timer function, and CLKGEN (clock generator, 618) indicates a clock. Represents a module that performs a generation function. The CLKGEN 618 injects a clock signal provided from inside or outside the apparatus shown in FIG. 4 to generate a clock signal provided to each of the component modules shown in FIG. 4, and in particular, adjusts the clock speed for low power consumption. do.

Similarly, PMEM (program memory, 620), PMIF (program memory interface, 622), EXIF (external interface, 624), MEMIF (memory interface, 626), HMM (observation probability operation, 628), SIF (serial interface, 630) ), UART (Asynchronous Serial Interface, 632), GPIO (Universal Interface, 634), CODEC IF (Codec Interface, 636), and CODEC (Codec, 640) are the program memory, program memory interface, external interface, memory Modules that perform interfaces, hidden Markov model operations, synchronous serial interfaces, asynchronous serial interfaces, general-purpose input / output, codec interfaces, and codec functions. In particular, the HMM 628 performs a word search operation using a predetermined hidden Markov model from these feature values.

In addition, the external bus 452 is an external bus for a data interface with an external memory. EXIF 624 supports Dynamic Memory Access (DMA). In particular, HMM 628 includes the apparatus of FIG. 3 for computing observation probabilities.

A controller (decoder, not shown) inside each component receives a command through an instruction bus (OPcode bus, 648, 650), decodes and performs a necessary operation. That is, the controller inside the HMM 628 receives and decodes a command through a control command bus (OPcode bus0, 1) to control the observation probability calculating apparatus of the present invention as shown in FIG. . On the other hand, the data are provided via two read buses 642 and 644 or output through one write bus 646.

The apparatus illustrated in FIG. 6 includes a program memory (PMEM, Program Memory, 620), and the program is loaded into the program memory (PMEM, 620) through an external interface device (EXIF, EXTERNAL InterFace, 624).

The HMM 628 receives a control command provided from a control unit (Ctrl Unit 602) illustrated in FIG. 6 through two OPcode buses 648 and 650, and a control command received from an internal control unit (not shown). Decode and control the observation probability computing device as shown in Figure 3 to perform the observation probability calculation.

FIG. 7 is a block diagram schematically illustrating a process of receiving a control command and data in the apparatus shown in FIG. 6.

The controller 602 controls to directly decode a control command to perform a specified operation, or controls the operation of each component module using the OPcode bus 0,1 (648, 650). Each configuration module shares OPcode bus1,2 and read bus A, B.

In case of direct control by the control unit (Ctrl Unit 602), the control command is fetched from the program memory (PMEM) 620, decoded, and the operands necessary for the control operation are read. It is stored in a register file (REG FILE, 604). After the control operation is control logic, ALU (Arithmatic Logic Unit, 606), multiplication and accumulation (MAC, Multiply and ACcummulate, 608), and barrel shift operation, B SHIFTER (Barrel SHITER, 610) In the case of square / root operation, the control operation is performed using SQRT (SQart and RooT, 614), and the result is stored in the register file (REG FILE, 604).

If the control unit (Ctrl Unit 602) does not control directly, the OPcode bus 0,1 (648, 650) is used. The control unit Ctrl unit 602 applies the control command patched in turn to the OPcode bus0 648 and the OPcode bus1 650 instead of decoding the control command captured from the program memory PMEM 620.

The same control command is applied to the OPcode bus0 648 and the OPcode bus1 650 one by one with a difference of one clock. When the control command is applied to the OPcode bus0 648, the configuration modules determine whether the control command corresponds to the control command. To this end, the configuration modules have decoders for decoding the control command. When the same control command is applied to the OPcode bus1 650 after one clock, control is performed to perform an operation corresponding to the designated control command at this time. RT and ET signal lines are allocated to indicate whether the control code applied to each of the OPcode buses 648 and 650 is enabled.

FIG. 8 is a timing diagram schematically showing a process of receiving a control command and data in the apparatus shown in FIG. 6.

In FIG. 8, the highest signal is a clock signal CLK, which in turn is a control command applied to OPcode bus0, a control command applied to OPcode bus1, an RT signal, an ET signal, data applied to read bus A, and read bus B. Data applied to.

When a control command is applied to the OPcode bus0 648 and enabled by the RT signal, one of the configuration modules of FIG. 4 recognizes it, decodes it, and enters a standby state. After the same control command is applied to the OPcode bus1 650 and enabled by the ET signal, the corresponding configuration module performs the operation specified by the control command. Specifically, data applied to the read bus A and the read bus B is received, a specified operation is performed, and a result value is output through the write bus.

According to the observation probability computing device of the present invention, in using the hidden Markov model search method, it is possible to efficiently perform the observation probability calculation that performs the most calculations.

Observation probability calculation dedicated device for searching hidden Markov model implemented through the present invention is invented to improve the processing speed of the speech recognition function, and can reduce the number of instructions more than 50% than without using such a device. Thus, if the same function is processed at a certain time, it can be processed at a lower clock speed and power consumption can be reduced to 1/2.

In addition, the apparatus can be used for stochastic calculations using hidden Markov models.

Claims (6)

A storage device for storing a mean of a parameter extracted from representative phonemes and a degree of distribution of the mean value (1 / σ); A subtractor for calculating a difference between an average provided from the storage device and a feature extracted from a speech signal to be recognized; And And a multiplier for multiplying the output of the subtractor and the degree of distribution provided from the storage device. The storage device of claim 1, wherein when i is an argument indicating a representative type of phoneme and j is an argument indicating a number of parameters, the storage device stores precision [i] [j] and mean [i] [j]. Providing them to the subtractor in a predetermined order; The subtractor subtracts the difference between mean [i] [j] and feature [i] [j] in the predetermined order, And the multiplier multiplies precision [i] [j] and a result of the subtraction of the subtractor according to the predetermined order. The observation probability calculating device according to claim 2, further comprising a square generator which squares the multiplication result of the multiplier. 4. The apparatus of claim 3, further comprising registers for buffering the precision [i] [j], mean [i] [j], and feature [i] [j], respectively. 4. The apparatus of claim 3, further comprising an accumulator for accumulating the output of the power supply. 6. The apparatus of claim 5, further comprising a register for buffering an accumulation result of the accumulator.