The invention relates to speech recognition. More particularly, the invention relates to efficient empirical determination, computation, and use of an acoustic confusability measure.
In United States Patent Application Publication No. 20020032549, it is stated:
In the operation of a speech recognition system, some acoustic information is acquired, and the system determines a word or word sequence that corresponds to the acoustic information. The acoustic information is generally some representation of a speech signal, such as the variations in voltage generated by a microphone. The output of the system is the best guess that the system has of the text corresponding to the given utterance, according to its principles of operation.
The principles applied to determine the best guess are those of probability theory. Specifically, the system produces as output the most likely word or word sequence corresponding to the given acoustic signal. Here, “most likely” is determined relative to two probability models embedded in the system: an acoustic model and a language model. Thus, if A represents the acoustic information acquired by the system, and W represents a guess at the word sequence corresponding to this acoustic information, then the system's best guess W* at the true word sequence is given by the solution of the following equation:
W*=argmaxWP(A|W)P(W).
Here P(A|W) is a number determined by the acoustic model for the system, and P(W) is a number determined by the language model for the system. A general discussion of the nature of acoustic models and language models can be found in “Statistical Methods for Speech Recognition,” Jelinek, The MIT Press, Cambridge, Mass. 1999, the disclosure of which is incorporated herein by reference. This general approach to speech recognition is discussed in the paper by Bahl et al., “A Maximum Likelihood Approach to Continuous Speech Recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume PAMI-5, pp. 179-190, March 1983, the disclosure of which is incorporated herein by reference.
The acoustic and language models play a central role in the operation of a speech recognition system: the higher the quality of each model, the more accurate the recognition system. A frequently-used measure of quality of a language model is a statistic known as the perplexity, as discussed in section 8.3 of Jelinek. For clarity, this statistic will hereafter be referred to as “lexical perplexity.” It is a general operating assumption within the field that the lower the value of the lexical perplexity, on a given fixed test corpus of words, the better the quality of the language model.
However, experience shows that lexical perplexity can decrease while errors in decoding words increase. For instance, see Clarkson et al., “The Applicability of Adaptive Language Modeling for the Broadcast News Task,” Proceedings of the Fifth International Conference on Spoken Language Processing, Sydney, Australia, November 1998, the disclosure of which is incorporated herein by reference. Thus, lexical perplexity is actually a poor indicator of language model effectiveness.
Nevertheless, lexical perplexity continues to be used as the objective function for the training of language models, when such models are determined by varying the values of sets of adjustable parameters. What is needed is a better statistic for measuring the quality of language models, and hence for use as the objective function during training.
United States Patent Application Publication No. 20020032549 teaches an invention that attempts to solve these problems by:
Providing two statistics that are better than lexical perplexity for determining the quality of language models. These statistics, called acoustic perplexity and the synthetic acoustic word error rate (SAWER), in turn depend upon methods for computing the acoustic confusability of words. Some methods and apparatuses disclosed herein substitute models of acoustic data in place of real acoustic data in order to determine confusability.
In a first aspect of the invention taught in United States Patent Application Publication No. 20020032549, two word pronunciations l(w) and l(x) are chosen from all pronunciations of all words in fixed vocabulary V of the speech recognition system. It is the confusability of these pronunciations that is desired. To do so, an evaluation model (also called valuation model) of l(x) is created, a synthesizer model of l(x) is created, and a matrix is determined from the evaluation and synthesizer models. Each of the evaluation and synthesizer models is preferably a hidden Markov model. The synthesizer model preferably replaces real acoustic data. Once the matrix is determined, a confusability calculation may be performed. This confusability calculation is preferably performed by reducing an infinite series of multiplications and additions to a finite matrix inversion calculation. In this manner, an exact confusability calculation may be determined for the evaluation and synthesizer models.
In additional aspects of the invention taught in United States Patent Application Publication No. 20020032549, different methods are used to determine certain numerical quantities, defined below, called synthetic likelihoods. In other aspects of the invention, (i) the confusability may be normalized and smoothed to better deal with very small probabilities and the sharpness of the distribution, and (ii) methods are disclosed that increase the speed of performing the matrix inversion and the confusability calculation. Moreover, a method for caching and reusing computations for similar words is disclosed.
Such teachings are yet limited and subject to improvement.
There are three related elements to the invention herein:
Empirically Derived Acoustic Confusability Measures
The first element comprises a means for determining the acoustic confusability of any two textual phrases in a given language. Some specific advantages of the means presented here are:
The second element comprises computational techniques for efficiently applying the acoustic confusability scoring mechanism. Previous inventions have alluded to the use of acoustic confusability measures, but notably do not discuss practical aspects of applying them. In any real-world practical scheme, it is often required to estimate the mutual acoustic confusability of tens of thousands of distinct phrases. Without efficient means of computing the measure, such computations rapidly become impractical. In this patent, we teach means for efficient application of our acoustic confusability measure, allowing practical application to very large-scale problems.
Method for Using Acoustic Confusability Measures
The third element comprises a method for using acoustic confusability measures, derived by whatever means (thus, not limited to the measure disclosed here), to make principled choices about which specific phrases to make recognizable by a speech recognition application.
There are three related elements to the presently preferred embodiment of invention disclosed herein:
Empirically Derived Acoustic Confusability Measures
The first element comprises a means for determining the acoustic confusability of any two textual phrases in a given language. Some specific advantages of the means presented here are:
The second element comprises computational techniques for efficiently applying the acoustic confusability scoring mechanism. Previous inventions have alluded to the use of acoustic confusability measures, but notably do not discuss practical aspects of applying such mechanisms. In any real-world practical scheme, it is often required to estimate the mutual acoustic confusability of tens of thousands of distinct phrases. Without efficient means of computing the measure, such computations rapidly become impractical. In this patent, we teach means for efficient application of our acoustic confusability score, allowing practical application to very large-scale problems.
Method for Using Acoustic Confusability Measures
The third element comprises a method for using acoustic confusability measures, derived by whatever means (thus, not limited to the measure disclosed here), to make principled choices about which specific phrases to make recognizable by a speech recognition application.
Empirically Derived Acoustic Confusability Measure
The immediately following discussion explains how to derive and compute an empirically derived acoustic confusability measure. The discussion is divided into several subsections;
We first establish some notation and nomenclature. The symbol or expression being defined appears in the left hand column; the associated text explains its meaning or interpretation. Italicized English words, in the associated text, give the nomenclature we use to refer to the symbol and the concept.
We first present an outline of the method, then present a detailed explanation of how to apply the method.
Outline of Method
The method comprises two basic steps. The first step is corpus processing, in which the original corpus is passed through the automatic speech recognition system of interest. This step is non-iterative; that is, the corpus is processed just once by the recognition system. The second step is development of a family of phoneme confusability models. This step is iterative; that is, it involves repeated passes over the corpus, at each step delivering an improved family of confusability models.
Corpus Processing
We assume that we have at our disposal some large and representative set of utterances, in some given human language, with associated reliable transcriptions. We refer to this as the corpus. By an utterance we mean a sound recording, represented in some suitable computer-readable form. By transcription we mean a conventional textual representation of the utterance; by reliable we mean that the transcription may be regarded as accurate. We refer to these transcriptions as the truth, or the true transcriptions.
In this step, we pass the utterances through an automatic speech recognition system, one utterance at a time. For each utterance, the recognition system generates a decoding, in a form called a decoded frame sequence, and a confidence score. As defined above, a frame is a brief audio segment of the input utterance.
The decoded frame sequence comprises the recognizer's best guess, for each frame of the utterance, of the phoneme being enunciated, in that audio frame. As defined above, a phoneme is one of a finite number of basic sound units of a human language.
This decoded frame sequence is then transformed, by a process that we describe below, into a much shorter decoded phoneme sequence. The confidence score is a measure, determined by the recognition system, of the likelihood that the given decoding is correct.
We then inspect the true transcription of the input utterance, and by a process that we describe below, transform the true transcription (which is just regular text, in the language of interest) into a true phoneme sequence.
Thus for each utterance we have confidence score, and a pair of phoneme sequences: the decoded phoneme sequence, and the true phoneme sequence. We refer to this entire collection as the recognized corpus, and denote it as P.
The recognized corpus constitutes the output of the corpus processing step.
Iterative Development of Probability Model Family
From the preceding step, we have at our disposal the recognized corpus P, comprising a large number of pairs of phoneme sequences.
In this step, we iteratively develop a sequence of probability model families. That is, we repeatedly pass through the recognized corpus, analyzing each pair of phoneme sequences to collect information regarding the confusability of any two phonemes. At the end of each pass, we use the information just collected to generate an improved family of probability models. We repeat the procedure until there is no further change in the family of probability models, or the change becomes negligible.
It is important to understand that this step as a whole comprises repeated iterations. In the detailed discussion below, we describe a single iteration, and the criterion for declaring the step as a whole complete.
The output of this step is a family of probability models, which estimates the acoustic confusability of any two members of the augmented phoneme alphabet 0′. From these estimates, by another method that we explain, we may then derive the acoustic confusability measure that we seek.
We now provide detailed descriptions of the steps outlined above.
Corpus Processing
Let X={<u1, T1>, . . . , <uC, TC>} be the corpus, comprising C pairs of utterances and transcriptions. For each <u, T> pair in X:
The purpose of the phoneme map m is to reduce the effective size of the phoneme alphabet, by collapsing minor variants within the phoneme alphabet into a single phoneme. An example would be the mapping of the “p closure” phoneme, often denoted pcl, to the regular p phoneme. Another example would be splitting phoneme pairs, known as diphones, into separate phonemes. This operation can simplify the calculation, and avoids the problem of too finely subdividing the available statistical evidence, which can lead to unreliable estimates of phoneme confusability.
However, this operation may be skipped, or in what amounts to the same thing, the map m may be the identity map on the phoneme alphabet.
Note: it will be obvious to one skilled in the art, that by suitable modification the map m may function to expand rather than to reduce the phoneme alphabet, for instance by including left and/or right phonetic context in the output phoneme. This modification is also claimed as part of this invention.
Thus if
f′=r r r r eI eI z z z z
is the decoded frame sequence, comprising 10 frames, the result of coalescing f′ is the decoded phoneme sequence
d=r eI z.
Here and above, r, eI and z are all members of the phoneme alphabet Φ. This phoneme sequence corresponds to the regular English language word “raise.” Note that d has three elements, respectively d1=r, d2=eI, and d3=z.
We denote the coalescing operation by the letter g, and write d=g(f′) for the action described above.
As above, h, eI, z, and i: are all members of the phoneme alphabet Φ. Note that t has four elements, respectively t1=h, t2eI, t3=z, and t4=i:.
It should be noted that there may be more than one valid pronunciation for a transcription T There are a number of ways of dealing with this:
By applying these steps sequentially to each element of the corpus X, we obtain the recognized corpus P={<u1, d(u1), t(u1), s(u1)>, . . . , <uC, d(uC), t(uC), s(uC)>}, or more succinctly P={<u1, d1, t1, s1>, . . . , <uC, dC, tC, sC>}.
Iterative Development of Probability Model Family
We now give the algorithm for the iterative development of the required probability model family, Π={p(d|t)}.
Note that each p(m)(x|y) satisfies 0<p(m)(x|y)<1, and so each δm+1)(x|y)>0.
Step 6a: Consider the entry <u, d, t, s> of P, with decoded phoneme sequence d=d1d2 . . . dN, containing N phonemes, and true phoneme sequence t=t1t2 . . . tQ, containing Q phonemes. Construct a rectangular lattice of dimension (N+1) rows by (Q+1) columns, and with an arc from a node (i, j) to each of nodes (i+1, j), (i, j+1) and (i+1, j+1), when present in the lattice. (Note: “node (i, j)” refers to the node in row i, column j of the lattice.) The phrase “when present in the lattice” means that arcs are created only for nodes with coordinates that actually lie within the lattice. Thus, for a node in the rightmost column, with coordinates (i, Q), only the arc (i, Q)→(i+1, Q) is created.)
Step 6b: Label
An example of such a lattice appears, in various versions, in
Step 6c: The Bellman-Ford dynamic programming algorithm is a well-known method for finding the shortest path through a directed graphic with no negative cycles. We apply it here to find the shortest path from the source node, which we define as node (0, 0), to the terminal node, which we define as node (N, Q).
Because there is only a single arc incident on each of these nodes, the minimum costs are respectively 0+3=3 and 0+2=2. In each case, this quantity is determined as (minimum cost to reach the immediately preceding node)+(cost of traversing the arc from the immediately preceding node).
It is evident that the path from (0, 0) is the minimum cost path, and this is indicated in
By repeated application of this process, the minimum cost path from the source node to each node of the lattice may be determined.
Because the arc costs are guaranteed to be non-negative, it is evident to one skilled in the art that the same computation may be performed, at possibly lower computational cost, using Dijkstra's shortest path first algorithm. The improvement follows from the fact that only the minimum cost path from the source node to the terminal node is required, and so the algorithm may be halted as soon as this has been determined.
The output of this step is a sequence of arcs A=a1, a2, . . . , ak, in the lattice L, known to comprise the minimum cost path from the source node to the terminal node. We write l(a) for the phoneme pair x|y that labels the arc a.
Step 6d: For each arc ai in the minimum cost path A, labeled with phoneme pair x|y=l(ai), increment the counter c(x|y) by 1.
This completes the description of the method to construct an empirically derived acoustic confusability measure. The means of using the result of this algorithm to compute the acoustic confusability of two arbitrary phrases is described below.
N-Best Variant of the Method
An important variant of the just-described method to construct an empirically derived acoustic confusability measure, which can improve the accuracy of the resulting measure, is as follows.
It is well known to those skilled in the art that the output of a recognizer R (or RG, for a grammar-based recognition system), may comprise not a single decoding D, comprising a pair f, s, but a so-called “N-best list,” comprising a ranked series of alternate decodings, written f1, s1, f2, s2, . . . , fB, sB. In this section we explain a variant of the basic method described above, called the “N-Best Variant,” which makes use of this additional information. The N-best variant involves changes to both the corpus processing step, and the iterative development of probability model family step, as follows.
N-Best Variant Corpus Processing
In the N-best variant of corpus processing, for each utterance u, each entry fi(u), si(u) in the N-best list is treated as a separate decoding. All other actions, taken for a decoding of u, are then performed as before. The result is a larger recognized corpus P′.
N-Best Variant Iterative Development of Probability Model Family
In the N-best variant of iterative development of probability model family, there are two changes. First, the input is the larger recognized corpus, P′, developed as described immediately above. Second, in step 6d, as described above, when processing a given entry <u, d, t, s> of P′, each count c(x|y) is incremented by the value s, which is the confidence score of the given entry, rather than by 1.
The rest of the algorithm is unchanged.
In the preceding sections we described how to determine the desired probability model family Π={p(d|t)}. In this section we explain how to use H to compute the acoustic confusability of two arbitrary phrases w and v.
Specifically, we give algorithms for computing two quantities, both relating to acoustic confusability. The first is the raw phrase acoustic confusability r(v|w). This is a measure of the acoustic similarity of phrases v and w. The second is the grammar-relative confusion probability p(v|w, G). This is an estimate of the probability that a grammar-constrained recognizer RG returns the phrase v as the decoding, when the true phrase was w. Note that no reference is made to any specific pronunciation, in either quantity.
In both cases, we must come to grips with the fact that the phrases v and w may have multiple acceptable pronunciations. There are a variety of ways of dealing with this, all of which are claimed as part of this patent.
In the process of computing these quantities, we also give expressions that depend upon specific pronunciations (and from which the pronunciation-free expressions are derived). These expressions have independent utility, and also are claimed as part of this patent.
Computation of Raw Pronunciation Acoustic Confusability r(q(v)|q(w)) and Raw Phrase Acoustic Confusability r(v|w)
We first assume that pronunciations q(w)∈Q(w) and q(v)∈Q(v) are given, and explain the computation of the raw pronunciation acoustic confusability, r(q(v)|q(w)). Then we explain methods to determine the raw phrase acoustic confusability r(v|w).
Computation of Raw Pronunciation Acoustic Confusability
Let the probability model family Π={p(d|t)} and the pronunciations q(w) and q(v) be given. Proceed as follows to compute the raw pronunciation acoustic confusability r(q(v)|q(w)):
Note that equivalently
and indeed this quantity may be computed directly from the lattice L, by suitable modification of the steps given above.
We have described here one method of computing a measure of the acoustic confusability r(q(v)|q(w)) of two pronunciations, q(w) and q(v). In what follows we describe methods of manipulating this measure to obtain other useful expressions. It is to be noted that while the expressions developed below assume the existence of some automatic means of quantitatively expressing the confusability of two pronunciations, they do not depend on the exact formulation presented here, and stand as independent inventions.
Computation of Raw Phrase Acoustic Confusability
We begin by defining r(v|q(w))=Σr(q(v)|q(w)), where the sum proceeds over all q(v)∈Q(v). This accepts any pronunciation q(v) as a decoding of v. The raw phrase acoustic confusability r(v|w), with no reference to pronunciations, may then be determined by any of the following means:
Those skilled in the art will observe ways to combine these methods into additional hybrid variants, for instance by randomly selecting q(v), but using the most common pronunciation q*(w), and setting r(v|w)=r(q(v) q*(w)).
Computation of Grammar-Relative Pronunciation Confusion Probability p(q(v)|q(w), G) and Grammar-Relative Phrase Confusion Probability p(v|w, G)
Suppose that a recognizer is constrained to recognize phrases within a grammar G. We proceed to define expressions that estimate the grammar-relative pronunciation confusion probability p(q(v)|q(w), G), and the grammar-relative phrase confusion probability p(v|w, G).
In what follows we write L(G) for the set of all phrases admissible by the grammar G, and Q(L(G)) for the set of all pronunciations of all such phrases. By assumption L(G) and Q(L(G)) are both finite.
Computation of Grammar Relative Pronunciation Confusion Probability p(q(v)|q(w), G)
Let two pronunciations q(v), q(w)∈Q(L(G)) be given; exact homonyms, that is q(v)=q(w), are to be excluded. We estimate p(q(v)|q(w), G), the probability that an utterance corresponding to the pronunciation q(w) is decoded by the recognizer RG as q(v), as follows.
Let two phrases v, w ∈L(G) be given. We estimate p(v|w, G), the probability that an utterance corresponding to any pronunciation of w is decoded by the recognizer RG as any pronunciation of v, as follows.
As above we must deal with the fact that there are in general multiple pronunciations of each phrase. We proceed in a similar manner, and begin by defining p(v|q(w),G)=Σp(q(v)|q(w),G), where the sum is taken over all q(v)∈Q(v). We may then proceed by one of the following methods:
In applying measures of acoustic confusability, it is typically necessary to compute a very large number of grammar-relative pronunciation confusion probabilities, p(q(v)|q(w), G), which ultimately depend upon the quantities r(q(v)|q(w)) and Z(q(w), G). We now explain three methods for improving the efficiency of these computations.
Partial Lattice Reuse
For a fixed q(w) in Q(L(G)), it is typically necessary to compute a large number of raw pronunciation confusability values r(q(v)|q(w)), as q(v) takes on each or many values of Q(L(G)). In principle for each q(v) this requires the construction, labeling and minimum-cost-path computation for the lattice L=q(v)×q(w), and this is prohibitively expensive.
This computation can be conducted more efficiently by exploiting the following observation. Consider two pronunciations q(v1)=d11, d12, . . . , d1Q1 and q(v2)=d21, d22, . . . , d2Q2. Suppose that they share a common prefix; that is, for some M≤Q1, Q2 we have d1j=d2j for j=1, . . . , M. Then the first M rows of the labeled and minimum-cost-path-marked lattice L1=q(v1)×q(w) can be reused in the construction, labeling and minimum-cost-path computation for lattice L2=q(v2)×q(w).
The reuse process consists of retaining the first (M+1) rows of nodes of the L1 lattice, and their associated arcs, labels and minimum-cost-path computation results, and then extending this to the L2 lattice, by adjoining nodes, and associated arcs and labels, corresponding to the remaining Q2-M phonemes of q(v2). Thereafter, the computation of the required minimum-cost-path costs and arcs proceeds only over the newly-added Q2-M bottom rows of L2.
For instance, continuing the exemplary lattice illustrated earlier, suppose q(w)=h eI z i:, and take q(v1)=r eI z (a pronunciation of “raise”) and q(v2)=r eI t (a pronunciation of “rate”). Then to transform L1=q(v1)×q(w) into L2=q(v2)×q(w) we first remove all the bottom row of nodes (those with row index of 3), and all arcs incident upon them. These all correspond to the phoneme “z” in q(v1). (However, we retain all other nodes, and all labels, values and computational results that mark them.) Then we adjoin a new bottom row of nodes, and associated arcs, all corresponding to the phoneme “t” in q(v2).
Note that it is possible, for example if q(v2)=r eI (a pronunciation of “ray”), that no additional nodes need be added, to transform L1 into L2. Likewise, if for example q(v2)=r eI {circle around (a)} r (a pronunciation of “razor”), it is possible that no nodes need to be removed.
This procedure may be codified as follows:
It will be obvious to one skilled in the art that this same technique may be applied, with appropriate modifications to operate on the columns rather than the rows of the lattice in question, by keeping q(v) fixed, and operating over an enumeration q(w1), q(w2), . . . of Q(L(G)) to compute a sequence of values r(q(v)|q(w1)), r(q(v)|q(w2)), . . . .
Pruning
One application of acoustic confusability measures is to find phrases within a grammar, vocabulary or phrase list that are likely to be confused. That is, we seek pairs of pronunciations q(v), q(w), both drawn from Q(L(G)), with v≠w, such that r(q(v)|q(w)), and hence ultimately p(q(v)|q(w), G), is large.
In principle, this involves the computation of r(q(v)|q(w)) for some |Q(L(G))|2 distinct pronunciation pairs. Because it is not uncommon for Q(L(G)) to contain as many as 100,000 members, this would entail on the order of 10 billion acoustic confusability computations. Because of the complexity of the computation, this is a daunting task for even a very fast computer.
However, it is possible to simplify this computation, as follows. If it can be established, with a small computational effort, that r(q(v)|q(w))<<r(q(w)|q(w)), then the expensive exact computation of r(q(v)|q(w)) need not be attempted. In this case we declare q(v) “not confusable” with q(w), and take r(q(v)|q(w))=0 in any further computations.
We refer to such a strategy as “pruning.” We now describe two complementary methods of pruning, respectively the method of Pronunciation Lengths, and the method of Pronunciation Sequences.
Pronunciation Lengths
Consider pronunciations q(v)=d1, d2, . . . , dD and q(w)=t1, t2, . . . , tT. Suppose for a moment that D>>T; in other words that q(v) contains many more phonemes than q(w). Then the minimum cost path through the lattice L=q(v)×q(w) necessarily traverses many edges labeled with insertion costs δ(x|ε), for some x in the phoneme sequence q(v). This entails a lower bound on the minimum cost path through L, which in turn entails an upper bound on r(q(v)|q(w)).
We now explain the method in detail. Let q(v)=d1, d2, . . . , dD and q(w)=t1, t2, . . . , tT, and let a threshold Θ be given. (The value of Θ may be a fixed number, a function of r(q(w)|q(w)), or determined in some other way.) We proceed to compute an upper bound r†(q(v)|q(w)) on r(q(v)|q(w)).
Let us write δi=δ(di|ε) for each phoneme di of q(v), where i=1, . . . , D. Sort these costs in increasing order, obtaining a sequence δi
Now, because D is the number of phonemes in q(v), even if the T phonemes of q(w) are exactly matched in the minimum cost path through the lattice, that path must still traverse at least I=−T arcs labeled with the insertion cost of some phoneme d of q(v). In other words, the cost S of the minimum cost path through the lattice is bounded below by the sum of the I smallest insertion costs listed above, S†=δi
Note: the computation of the exponential can be avoided if we take B=log Θ, and equivalently check that −B≤S†.
A similar bound may be developed for the case T>>D. For this case we consider the phoneme deletion costs δi=δ(ε|ti) for each phoneme ti of q(w), where i=1, . . . , T. As before, we sort these costs, obtaining the sequence δi
Pronunciation Sequences
The preceding method of Pronunciation Lengths required either D>>T or T>>D, where these are the lengths of the respective pronunciation sequences. We now describe a method that may be applied, under suitable conditions, when DT
For each ϕ in Φ, define δsdmin(ϕ)=min{δ(x|ϕ)|x∈Φ′}, and define δsimin(ϕ)=min{δ(ϕ|x∈Φ′}. Thus δsdmin(ϕ) is the minimum of all costs to delete ϕ or substitute any other phoneme for ϕ, and likewise δsimin(ϕ) is the minimum of all costs to insert ϕ or substitute ϕ for any other phoneme. Note that these values are independent of any particular q(v) and q(w), and may be computed once for all time.
To apply the method, as above let q(v)=d1, d2 . . . , dD and q(w)=t1, t2, . . . , tT, and let a threshold Θ be given.
For each ϕ in Φ, define w #(ϕ) and v #(ϕ) to be the number of times the phoneme ϕ appears in q(w) and q(v) respectively. Let n(ϕ)=w #(ϕ)−v #(ϕ).
Now form the sequence W\V=ϕ1, ϕw, . . . , where for each ϕ in Φ with n(ϕ)>0, we insert n(ϕ) copies of ϕ into the sequence. Note that a given ϕ may occur multiple times in W\V, and observe that for each instance of ϕ in W\V, the minimum cost path through the lattice L=q(v)×q(w) must traverse a substitution or deletion arc for ϕ.
Now compute S†=Σδsdmin(ϕ), where the sum runs over the entries of W\V. It follows that S, the cost of the true minimal cost path through L, is bounded below by S†. Hence we may define r†(q(v)|q(w))=exp(−S†) and proceed as before.
A similar method applies with the sequence V\W, where we insert n(ϕ)=v #(ϕ)−w #(ϕ) copies of ϕ in the sequence, for n(ϕ)>0. (Note the interchange of v and w here.) We compute S†=Σδsimin(ϕ), where the sum runs over the entries of V\W, and proceed as above.
Incremental Computation of Confusability in a Sequence of Grammars
Suppose have two grammars, G and G′, such that L(G) and L(G′) differ from one another by a relatively small number of phrases, and hence so that Q(L(G)) and Q(L(G′)) differ by only a small number of pronunciations. Let us write Q and Q′ for these two pronunciation lists, respectively.
Suppose further that we have already computed a full set of grammar-relative pronunciation confusion probabilities, p(q(v) q(w), G), for the grammar G. Then we may efficiently compute a revised set p(q(v) q(w), G′), as follows.
First observe that the value of a raw pronunciation confusion measure, r(q(v) q(w)), is independent of any particular grammar. While Q′ may contain some pronunciations not in Q, for which new values r(q(v) q(w)) must be computed, most will already be known. We may therefore proceed as follows.
We now present two of the primary applications of an acoustic confusability measure.
The first of these, the “Confusatron,” is a computer program that takes as input an arbitrary grammar G, with a finite language L(G), and finds phrases in L(G) that are likely to be frequent sources of error, for the speech recognition system. The second is a method, called maximum utility grammar augmentation, for deciding in a principled way whether or not to add a particular phrase to a grammar.
While our discussion presumes the existence of a raw pronunciation confusability measure r(q(v)|q(w)), and/or grammar-relative pronunciation confusion probabilities p(q(v)|q(w), G), the methods presented in this section are independent of the particular measures and probabilities developed in this patent, and stand as independent inventions.
The Confusatron
We now explain a computer program, which we refer to as the “Confusatron,” which automatically analyzes a given grammar G to find so-called “dangerous words.” These are actually elements of L(G) with pronunciations that are easily confusable, by a given automatic speech recognition technology.
The value of the Confusatron is in its ability to guide a speech recognition system designer to decide what phrases are recognized with high accuracy within a given application, and which are not. If a phrase identified as likely to be poorly recognized may be discarded and replaced by another less confusable one, in the design phase, the system is less error-prone, and easier to use. If a phrase is likely to be troublesome, but must nevertheless be included in the system, the designer is at least forewarned, and may attempt to take some mitigating action.
We begin with a description of the Confusatron's function, and its basic mode of operation. We then describe variations; all are claimed as part of the patent.
The Confusatron generates a printed report, comprising two parts.
The first part, an example of which is exhibited in
However, it is the second part that is really useful. Here the Confusatron automatically identifies words with distinct pronunciations that are nevertheless likely to be confused. This is the “dangerous word” list, an example of which is exhibited in
The Confusatron operates as follows. Let G be a grammar, with finite language L(G), and finite pronunciation set Q(L(G)). Let {p(q(v)|q(w), G)} be a family of grammar-relative pronunciation confusability models, either derived from an underlying raw pronunciation confusion measure r(q(v)|q(w)) as described above, or defined by independent means.
It is useful at this point to introduce the quantity C(q(w), G), called the “clarity” of q(w) in G. This is a statistic of our invention, which is defined by the formula
The unit of this statistic, defined as above, is called a “deciclar,” where “clar” is pronounced to rhyme with “car.” This turns out to be a convenient expression, and unit, in which to measure the predicted recognizability of a given pronunciation q(w), within a given grammar G. Note that the clarity is defined with reference to a particular grammar. If the grammar is clear from context, we do not mention it or denote it in symbols.
Note that the higher the value of p(q(w)|q(w), G), which is the estimated probability that q(w) is recognized as itself, when enunciated by a competent speaker, the larger the value of C(q(w), G). Thus high clarity pronunciations are likely to be correctly decoded, whereas lower clarity pronunciations are less likely to be correctly decoded. This forms the basic operating principle of the Confusatron, which we now state in detail.
Several important variations of the basic Confusatron algorithm are now noted.
Results for Pronunciations
First, rather than aggregating and presenting clarity results C(q(w), G) over all q(w) in Q(w), it is sometimes preferable to report them for individual pronunciations q(w). This can be useful if it is desirable to identify particular troublesome pronunciations.
Semantic Fusion
Second, there is often some semantic label attached to distinct phrases v and w in a grammar, such that they are known to have the same meaning. If they also have similar pronunciations (say, they differ by the presence of some small word, such as “a”), it is possible that the value of p(q(v)|q(w), G) is high. This may nominally cause q(w) to have low clarity, and thereby lead to flagging w as dangerous, when in fact the pronunciations q(v) that are confusable with q(w) have same underlying meaning to the speech recognition application.
It is straightforward to analyze the grammar's semantic labels, when they are present, and accumulate the probability mass of each p(q(v)|q(w), G) into p(q(w)|q(w), G), in those cases when v and w have the same meaning. This process is known as “semantic fusion,” and it is a valuable improvement on the basic Confusatron, which is also claimed in this patent.
Dangerous Word Detection Only
Suppose our task is only to decide if a given pronunciation q(w) is dangerous or not, that is if C(q(w), G)<Γ. By straightforward algebra, this can be turned into an equivalent comparison p(q(w)|q(w),G)<10(Γ/10)/(1+10(Γ/10)). Let us write Ψ for this transformed threshold 10(Γ/10)/(1+10(Γ/10).
Recall that p(q(w)|q(w), G)=r(q(w) q(w))|Z(q(w), G), and that the denominator is a monotonically growing quantity, as the defining sum proceeds over all q(v) in Q(L(G)), excluding homonyms of q(w). Now by definition p(q(w)|q(w), G)<Ψiff r(q(w)|q(w))/Z(q(w), G)<Ψ, that is iff Z(q(w), G)>r(q(w)|q(w))/Ψ.
Thus, we can proceed by first computing r(q(w)|q(w)), then accumulating Z(q(w), G), which is defined as Z(q(w),G)=Σr(q(x)|q(w)), where the sum runs over all non-homonyms of q(w) in Q(L(G)), and stopping as soon as the sum exceeds r(q(w)|q(w))/Ψ. If we arrange to accumulate into the sum the quantities r(q(x)|q(w)) that we expect to be large, say by concentrating on pronunciations of length close to that of q(w), then for dangerous words we may hope to terminate the accumulation of Z(q(w), G) without proceeding all the way through Q(L(G)).
Maximum Utility Grammar Augmentation
Suppose we are given a predetermined utility U(w) for recognizing a phrase w in a speech recognition application, a prior probability p(w) of the phrase. Then we may define the value of the phrase, within a grammar G, as V(w, G)=p(w) p(w|w, G) U(w). We may then further define the value of a grammar as the value of all its recognizable phrases; that is, V(G)=ΣV(w,G), where the sum extends over all w in L(G).
Consider now some phrase w that is not in L(G); we are trying to decide whether to add it to G or not. On the one hand, presumably adding the phrase has some value, in terms of enabling new functionality for a given speech recognition application, such as permitting the search, by voice, for a given artist or title in a content catalog.
On the other hand, adding the phrase might also have some negative impact, if it has pronunciations that are close to those of phrases already in the grammar: adding the new phrase could induce misrecognition of the acoustically close, already-present phrases.
Let us write G+w for the grammar G with w added to it. Then a principled way to decide whether or not to add a given phrase w is to compute the gain in value ΔV(w), defined as ΔV(w)=V(G+w)−V(G).
Moreover, given a list of phrases w1, w2, . . . , under consideration for addition to G, this method can be used to rank their importance, by considering each ΔV(w i), and adding the phrases in a greedy manner. By recomputing the value gains at each stage, and stopping when the value gain is no longer positive, a designer can be assured of not inducing any loss in value, by adding too many new phrases.
Although the invention is described herein with reference to the preferred embodiment, one skilled in the art will readily appreciate that other applications may be substituted for those set forth herein without departing from the spirit and scope of the present invention. Accordingly, the invention should only be limited by the Claims included below.
This application is a continuation of U.S. patent application Ser. No. 16/988,292, filed Aug. 7, 2020, which is a continuation of U.S. patent application Ser. No. 16/158,900, filed Oct. 12, 2018, now U.S. Pat. No. 10,748,527, issued Aug. 18, 2020, which is a continuation of U.S. patent application Ser. No. 15/457,964, filed Mar. 13, 2017, now U.S. Pat. No. 10,121,469, issued Nov. 6, 2018, which is a divisional of U.S. patent application Ser. No. 14/574,314, filed Dec. 17, 2014, now U.S. Pat. No. 9,626,965, issued Apr. 18, 2017, which is a divisional application of U.S. patent application Ser. No. 11/932,122, filed Oct. 31, 2007, now U.S. Pat. No. 8,959,019, issued Feb. 17, 2015, which are incorporated herein in their entireties by this reference thereto.
Number | Name | Date | Kind |
---|---|---|---|
4980918 | Bahl et al. | Dec 1990 | A |
5381459 | Lappington | Jan 1995 | A |
5553119 | McAllister et al. | Sep 1996 | A |
5581655 | Cohen et al. | Dec 1996 | A |
5611019 | Nakatoh et al. | Mar 1997 | A |
5698834 | Worthington et al. | Dec 1997 | A |
5737723 | Riley et al. | Apr 1998 | A |
5752232 | Basore et al. | May 1998 | A |
5754977 | Gardner et al. | May 1998 | A |
5774859 | Houser et al. | Jun 1998 | A |
5963903 | Hon et al. | Oct 1999 | A |
5987411 | Petroni et al. | Nov 1999 | A |
6009387 | Ramaswamy et al. | Dec 1999 | A |
6012058 | Fayyad et al. | Jan 2000 | A |
6021387 | Mozer et al. | Feb 2000 | A |
6049767 | Printz | Apr 2000 | A |
6073099 | Sabourin et al. | Jun 2000 | A |
6130726 | Darbee et al. | Oct 2000 | A |
6134527 | Meunier et al. | Oct 2000 | A |
6141640 | Moo | Oct 2000 | A |
6182039 | Rigazio et al. | Jan 2001 | B1 |
6185530 | Ittycheriah et al. | Feb 2001 | B1 |
6195641 | Loring et al. | Feb 2001 | B1 |
6243679 | Mohri et al. | Jun 2001 | B1 |
6260013 | Sejnoha | Jul 2001 | B1 |
6263308 | Heckerman et al. | Jul 2001 | B1 |
6298324 | Zuberec et al. | Oct 2001 | B1 |
6301560 | Masters | Oct 2001 | B1 |
6320947 | Joyce et al. | Nov 2001 | B1 |
6336091 | Polikaitis et al. | Jan 2002 | B1 |
6374177 | Lee et al. | Apr 2002 | B1 |
6374226 | Hunt et al. | Apr 2002 | B1 |
6381316 | Joyce et al. | Apr 2002 | B2 |
6408272 | White et al. | Jun 2002 | B1 |
6415257 | Junqua et al. | Jul 2002 | B1 |
6424935 | Taylor | Jul 2002 | B1 |
6446035 | Grefenstette et al. | Sep 2002 | B1 |
6493667 | De Souza et al. | Dec 2002 | B1 |
6523005 | Holzapfel | Feb 2003 | B2 |
6658414 | Bryan et al. | Dec 2003 | B2 |
6665644 | Kanevsky et al. | Dec 2003 | B1 |
6711541 | Kuhn et al. | Mar 2004 | B1 |
6711543 | Cameron | Mar 2004 | B2 |
6714632 | Joyce et al. | Mar 2004 | B2 |
6721633 | Funk et al. | Apr 2004 | B2 |
6725022 | Clayton et al. | Apr 2004 | B1 |
6728531 | Lee et al. | Apr 2004 | B1 |
6754625 | Olsen et al. | Jun 2004 | B2 |
6799201 | Lee et al. | Sep 2004 | B1 |
6804653 | Gabel | Oct 2004 | B2 |
6892083 | Shostak | May 2005 | B2 |
6901366 | Kuhn et al. | May 2005 | B1 |
6975993 | Keiller | Dec 2005 | B1 |
6985865 | Packingham et al. | Jan 2006 | B1 |
7013276 | Bickley et al. | Mar 2006 | B2 |
7020609 | Thrift et al. | Mar 2006 | B2 |
7027987 | Franz et al. | Apr 2006 | B1 |
7062477 | Fujiwara et al. | Jun 2006 | B2 |
7113981 | Slate | Sep 2006 | B2 |
7117159 | Packingham et al. | Oct 2006 | B1 |
7158959 | Chickering et al. | Jan 2007 | B1 |
7188066 | Falcon et al. | Mar 2007 | B2 |
7203645 | Pokhariyal et al. | Apr 2007 | B2 |
7219056 | Axelrod et al. | May 2007 | B2 |
7231380 | Pienkos | Jun 2007 | B1 |
7263487 | Hwang | Aug 2007 | B2 |
7263489 | Cohen et al. | Aug 2007 | B2 |
7277851 | Henton | Oct 2007 | B1 |
7310600 | Garner et al. | Dec 2007 | B1 |
7324947 | Jordan et al. | Jan 2008 | B2 |
7406417 | Hain | Jul 2008 | B1 |
7428555 | Yan | Sep 2008 | B2 |
7444282 | Choo et al. | Oct 2008 | B2 |
7447636 | Schwartz et al. | Nov 2008 | B1 |
7483885 | Chandrasekar et al. | Jan 2009 | B2 |
7519534 | Maddux et al. | Apr 2009 | B2 |
7590605 | Josifovski | Sep 2009 | B2 |
7654455 | Bhatti et al. | Feb 2010 | B1 |
7769786 | Patel | Aug 2010 | B2 |
7809601 | Shaya et al. | Oct 2010 | B2 |
7831549 | Tilei et al. | Nov 2010 | B2 |
7844456 | Cai et al. | Nov 2010 | B2 |
7860716 | Tian et al. | Dec 2010 | B2 |
7881930 | Faisman et al. | Feb 2011 | B2 |
7904296 | Morris | Mar 2011 | B2 |
7934658 | Bhatti et al. | May 2011 | B1 |
7949526 | Ju et al. | May 2011 | B2 |
7974843 | Schneider | Jul 2011 | B2 |
8165916 | Hoffberg et al. | Apr 2012 | B2 |
8306818 | Chelba et al. | Nov 2012 | B2 |
8321278 | Haveliwala et al. | Nov 2012 | B2 |
8321427 | Stampleman et al. | Nov 2012 | B2 |
8374870 | Braho et al. | Feb 2013 | B2 |
8515753 | Kim et al. | Aug 2013 | B2 |
8577681 | Roth et al. | Nov 2013 | B2 |
8793127 | Printz et al. | Jul 2014 | B2 |
8959019 | Printz et al. | Feb 2015 | B2 |
9626965 | Printz et al. | Apr 2017 | B2 |
10121469 | Printz et al. | Nov 2018 | B2 |
10748527 | Printz et al. | Aug 2020 | B2 |
11587558 | Printz | Feb 2023 | B2 |
20010019604 | Joyce et al. | Sep 2001 | A1 |
20010037324 | Agrawal et al. | Nov 2001 | A1 |
20020015480 | Daswani et al. | Feb 2002 | A1 |
20020032549 | Axelrod et al. | Mar 2002 | A1 |
20020032564 | Ehsani et al. | Mar 2002 | A1 |
20020044226 | Risi | Apr 2002 | A1 |
20020046030 | Haritsa et al. | Apr 2002 | A1 |
20020049535 | Rigo et al. | Apr 2002 | A1 |
20020075249 | Kubota et al. | Jun 2002 | A1 |
20020106065 | Joyce et al. | Aug 2002 | A1 |
20020107695 | Roth et al. | Aug 2002 | A1 |
20020116190 | Rockenbeck et al. | Aug 2002 | A1 |
20020116191 | Olsen et al. | Aug 2002 | A1 |
20020133340 | Basson et al. | Sep 2002 | A1 |
20020146015 | Bryan et al. | Oct 2002 | A1 |
20030004728 | Keiller | Jan 2003 | A1 |
20030028380 | Freeland et al. | Feb 2003 | A1 |
20030033152 | Cameron | Feb 2003 | A1 |
20030046071 | Wyman | Mar 2003 | A1 |
20030061039 | Levin | Mar 2003 | A1 |
20030065427 | Funk et al. | Apr 2003 | A1 |
20030068154 | Zylka | Apr 2003 | A1 |
20030069729 | Bickley et al. | Apr 2003 | A1 |
20030073434 | Shostak | Apr 2003 | A1 |
20030088416 | Griniasty | May 2003 | A1 |
20030093281 | Geilhufe et al. | May 2003 | A1 |
20030125928 | Lee et al. | Jul 2003 | A1 |
20030177013 | Falcon et al. | Sep 2003 | A1 |
20030212702 | Campos et al. | Nov 2003 | A1 |
20040039570 | Harengel et al. | Feb 2004 | A1 |
20040077334 | Joyce et al. | Apr 2004 | A1 |
20040110472 | Witkowski et al. | Jun 2004 | A1 |
20040127241 | Shostak | Jul 2004 | A1 |
20040132433 | Stern et al. | Jul 2004 | A1 |
20040153319 | Yacoub | Aug 2004 | A1 |
20040193408 | Hunt | Sep 2004 | A1 |
20040199498 | Kapur et al. | Oct 2004 | A1 |
20040249639 | Kammerer | Dec 2004 | A1 |
20050010412 | Aronowitz | Jan 2005 | A1 |
20050071224 | Fikes et al. | Mar 2005 | A1 |
20050125224 | Myers et al. | Jun 2005 | A1 |
20050143139 | Park et al. | Jun 2005 | A1 |
20050144251 | Slate | Jun 2005 | A1 |
20050170863 | Shostak | Aug 2005 | A1 |
20050182558 | Maruta | Aug 2005 | A1 |
20050198056 | Dumais et al. | Sep 2005 | A1 |
20050203751 | Stevens et al. | Sep 2005 | A1 |
20050228670 | Mahajan et al. | Oct 2005 | A1 |
20060018440 | Watkins et al. | Jan 2006 | A1 |
20060028337 | Li | Feb 2006 | A1 |
20060050686 | Velez-Rivera et al. | Mar 2006 | A1 |
20060064177 | Tian et al. | Mar 2006 | A1 |
20060074656 | Mathias et al. | Apr 2006 | A1 |
20060085521 | Sztybel | Apr 2006 | A1 |
20060136292 | Bhati et al. | Jun 2006 | A1 |
20060149635 | Bhatti et al. | Jul 2006 | A1 |
20060184365 | Odell et al. | Aug 2006 | A1 |
20060206339 | Silvera et al. | Sep 2006 | A1 |
20060206340 | Silvera et al. | Sep 2006 | A1 |
20060259467 | Westphal | Nov 2006 | A1 |
20060271546 | Phung | Nov 2006 | A1 |
20060287856 | He et al. | Dec 2006 | A1 |
20070027864 | Collins et al. | Feb 2007 | A1 |
20070033003 | Morris | Feb 2007 | A1 |
20070067285 | Blume et al. | Mar 2007 | A1 |
20070150275 | Garner et al. | Jun 2007 | A1 |
20070179784 | Thambiratnam et al. | Aug 2007 | A1 |
20070192309 | Fischer et al. | Aug 2007 | A1 |
20070198265 | Yao | Aug 2007 | A1 |
20070213979 | Meermeier | Sep 2007 | A1 |
20070214140 | Dom et al. | Sep 2007 | A1 |
20070219798 | Wang et al. | Sep 2007 | A1 |
20070250320 | Chengalvarayan | Oct 2007 | A1 |
20070271241 | Morris et al. | Nov 2007 | A1 |
20080021860 | Wiegering et al. | Jan 2008 | A1 |
20080046250 | Agapi et al. | Feb 2008 | A1 |
20080082322 | Joublin et al. | Apr 2008 | A1 |
20080103887 | Oldham et al. | May 2008 | A1 |
20080103907 | Maislos et al. | May 2008 | A1 |
20080126100 | Grost et al. | May 2008 | A1 |
20080154596 | Da Palma et al. | Jun 2008 | A1 |
20080250448 | Rowe et al. | Oct 2008 | A1 |
20090048910 | Shenfield et al. | Feb 2009 | A1 |
Number | Date | Country |
---|---|---|
0635820 | Jan 1995 | EP |
1341363 | Sep 2003 | EP |
1447792 | Aug 2004 | EP |
1003018 | May 2005 | EP |
1633150 | Mar 2006 | EP |
1633151 | Mar 2006 | EP |
1742437 | Jan 2007 | EP |
00016568 | Mar 2000 | WO |
00021232 | Apr 2000 | WO |
01022112 | Mar 2001 | WO |
01022249 | Mar 2001 | WO |
01022633 | Mar 2001 | WO |
01022712 | Mar 2001 | WO |
01022713 | Mar 2001 | WO |
01039178 | May 2001 | WO |
01057851 | Aug 2001 | WO |
02007050 | Jan 2002 | WO |
02011120 | Feb 2002 | WO |
02017090 | Feb 2002 | WO |
02097590 | Dec 2002 | WO |
04077721 | Sep 2004 | WO |
06033841 | Mar 2006 | WO |
06098789 | Sep 2006 | WO |
04021149 | Mar 2007 | WO |
05079254 | May 2007 | WO |
06029269 | May 2007 | WO |
Entry |
---|
“BBN Intros Speech Recognition for Cellular/Phone Apps”, Newsbytes, U.S.A., Feb. 28, 1995. |
“BBN's Voice Navigation for Time-Warner's FSN”, Telemedia News & Views, vol. 2, Issue 12, Dec. 1994, U.S.A. |
“Full Servce Network”, Time Warner Cable, The TWC Story | Eras Menu, 1990-1995, U.S.A. |
Colman, P., “The Power of Speech”, Convergence, Aug. 1995, pp. 16-23, U.S.A. |
Dawson, F., “Time Warner Pursues Voice as New Remote”, Broadband Week, Multichannel News, U.S.A., Jan. 1, 1995, pp. 31 and 34. |
Frozena, J., “(BBN) Time Warner Cable and BBN Hark Systems Corporation Plan to Provide Voice Access to the Information Superhighway”, Business Wire, Cambridge. Massachusetts, U.S.A., Nov. 1, 1994. |
Henriques, D., “Dragon Systems, a Former Little Guy, Gets Ready for Market”, New York Times, Business Day, Technology: Market Place, U.S.A., Mar. 1, 1999. |
Lefkowitz, L., “Voice-Recognition Home TV Coming This Year; Service Merges Computer, Phone, and Cable Technologies”, Computer Shopper, vol. 15, p. 68, Feb. 1995, U.S. and U.K. |
Wikipedia, “Additive smoothing”, 5 pages, downloaded Apr. 8, 2020. (Year: 2020). |
Wikipedia, “Law of succession”, 10 pages, downloaded Apr. 8, 2020. (Year: 2020). |
Salami , et al., “A Fully Vector Quantised Self-Excited Vocoder”, Int'l Conference on Acoustics, Speech & Signal Processing; vol. 1, par. 3.1; Glasgow, May 1989. |
Schotz, S. , “Automatic prediction of speaker age using CART”, Course paper for course in Speech Recognition, Lund University, retrieved online from url: http://person2.sol.lu.se/SusznneSchotz/downloads/SR_paper_SusanneS2004.pdf, 2003, 8 pages. |
“Dijkstra's Algorithm”, Wikipedia, downloaded Jul. 18, 2016., Jul. 8, 2016. |
Amir, A. , et al., “Advances in Phonetic Word Spotting”, IBM Research Report RJ 10215, Aug. 2001, pp. 1-3. |
Belzer , et al., “Symmetric Trellis-Coded Vector Quantization”, IEEE Transactions on Communications, IEEE Service Center, Piscataway, NJ, vol. 45, No. 45, par. II, figure 2, Nov. 1997, pp. 1354-1357. |
Chan , et al., “Efficient Codebook Search Procedure for Vector-Sum Excited Linear Predictive Coding of Speech”, IEEE Electronics Letters; vol. 30, No. 22; Stevanage, GB, ISSN 0013-5194, Oct. 27, 1994, pp. 1830-1831. |
Chan , “Fast Stochastic Codebook Search Through the Use of Odd-Symmetric Crosscorrelation Basis Vectors”, Int'l Conference on Acoustics, Speech and Signal Processing; Detroit, Michigan, vol. 1, Par. 1; ISBN 0-7803-2461-5, May 1995, pp. 21-24. |
Chen , et al., “Diagonal Axes method (DAM): A Fast Search Algorithm for Vecotr Quantization”, IEEE Transactions on Circuits and Systems for Video Technology, Piscataway, NJ; vol. 7, No. 3, ISSN 1051-8215; Par. I, II, Jun. 1997. |
Hanzo , et al., “Voice Compression and Communications—Principles and Applications for Fixed and Wireless Channels”, Wiley, ISBN 0-471-15039-8; par. 4.3.3, 2001. |
Number | Date | Country | |
---|---|---|---|
20230206914 A1 | Jun 2023 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 14574314 | Dec 2014 | US |
Child | 15457964 | US | |
Parent | 11932122 | Oct 2007 | US |
Child | 14574314 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 16988292 | Aug 2020 | US |
Child | 18171204 | US | |
Parent | 16158900 | Oct 2018 | US |
Child | 16988292 | US | |
Parent | 15457964 | Mar 2017 | US |
Child | 16158900 | US |