1、Speech RecognitionVictor Zue, Ron Cole, & Wayne WardMIT Laboratory for Computer Science, Cambridge, Massachusetts, USA Oregon Graduate Institute of Science & Technology, Portland, Oregon, USA Carnegie Mellon University Pittsburgh, Pennsylvania, USA1 Defining the ProblemSpeech recognition is the proc
2、ess of converting an acoustic signal, captured by a microphone or a telephone, to a set of words. The recognized words can be the final results, as for applications such as commands & control, data entry, and document preparation. They can also serve as the input to further linguistic processing in
3、order to achieve speech understanding, a subject covered in section.Speech recognition systems can be characterized by many parameters, some of the more important of which are shown in Figure. An isolated-word speech recognition system requires that the speaker pause briefly between words, whereas a
4、 continuous speech recognition system does not. Spontaneous, or extemporaneously generated, speech contains disfluencies, and is much more difficult to recognize than speech read from script. Some systems require speaker enrollmenta user must provide samples of his or her speech before using them, w
5、hereas other systems are said to be speaker-independent, in that no enrollment is necessary. Some of the other parameters depend on the specific task. Recognition is generally more difficult when vocabularies are large or have many similar-sounding words. When speech is produced in a sequence of wor
6、ds, language models or artificial grammars are used to restrict the combination of words.The simplest language model can be specified as a finite-state network, where the permissible words following each word are given explicitly. More general language models approximating natural language are speci
7、fied in terms of a context-sensitive grammar.One popular measure of the difficulty of the task, combining the vocabulary size and the 1 language model, is perplexity, loosely defined as the geometric mean of the number of words that can follow a word after the language model has been applied (see se
8、ction for a discussion of language modeling in general and perplexity in particular). Finally, there are some external parameters that can affect speech recognition system performance, including the characteristics of the environmental noise and the type and the placement of the microphone.Speech re
9、cognition is a difficult problem, largely because of the many sources of variability associated with the signal. First, the acoustic realizations of phonemes, the smallest sound units of which words are composed, are highly dependent on the context in which they appear. These phonetic variabilities
10、are exemplified by the acoustic differences of the phoneme, At word boundaries, contextual variations can be quite dramaticmaking gas shortage sound like gash shortage in American English, and devo andare sound like devandare in Italian.Second, acoustic variabilities can result from changes in the e
11、nvironment as well as in the position and characteristics of the transducer. Third, within-speaker variabilities can result from changes in the speakers physical and emotional state, speaking rate, or voice quality. Finally, differences in sociolinguistic background, dialect, and vocal tract size an
12、d shape can contribute to across-speaker variabilities.Figure shows the major components of a typical speech recognition system. The digitized speech signal is first transformed into a set of useful measurements or features at a fixed rate, 2 typically once every 1020 msec (see sectionsand 11.3 for
13、signal representation and digital signal processing, respectively). These measurements are then used to search for the most likely word candidate, making use of constraints imposed by the acoustic, lexical, and language models. Throughout this process, training data are used to determine the values
14、of the model parameters.Speech recognition systems attempt to model the sources of variability described above in several ways. At the level of signal representation, researchers have developed representations that emphasize perceptually important speaker-independent features of the signal, and de-e
15、mphasize speaker-dependent characteristics. At the acoustic phonetic level, speaker variability is typically modeled using statistical techniques applied to large amounts of data. Speaker adaptation algorithms have also been developed that adapt speaker-independent acoustic models to those of the cu
16、rrent speaker during system use, (see section). Effects of linguistic context at the acoustic phonetic level are typically handled by training separate models for phonemes in different contexts; this is called context dependent acoustic modeling.Word level variability can be handled by allowing alte
17、rnate pronunciations of words in representations known as pronunciation networks. Common alternate pronunciations of words, as well as effects of dialect and accent are handled by allowing search algorithms to find alternate paths of phonemes through these networks. Statistical language models, base
18、d on estimates of the frequency of occurrence of word sequences, are often used to guide the search through the most probable sequence of words.The dominant recognition paradigm in the past fifteen years is known as hidden Markov models (HMM). An HMM is a doubly stochastic model, in which the genera
19、tion of the underlying phoneme string and the frame-by-frame, surface acoustic realizations are both represented probabilistically as Markov processes, as discussed in sections,and 11.2. Neural networks have also been used to estimate the frame based scores; these scores are then integrated into HMM
20、-based system architectures, in what has come to be known as hybrid systems, as described in section 11.5.An interesting feature of frame-based HMM systems is that speech segments are identified during the search process, rather than explicitly. An alternate approach is to first identify speech segm
21、ents, then classify the segments and use the segment scores to recognize words. This approach has produced competitive recognition performance in several tasks.2 State of the ArtComments about the state-of-the-art need to be made in the context of specific applications which reflect the constraints
22、on the task. Moreover, different technologies are sometimes appropriate for different tasks. For example, when the vocabulary is small, the entire word can be modeled as a single unit. Such an approach is not practical for large vocabularies, where word models must be built up from subword units.The
23、 past decade has witnessed significant progress in speech recognition technology. Word error rates continue to drop by a factor of 2 every two years. Substantial progress has been made in the basic technology, leading to the lowering of barriers to speaker independence, continuous speech, and large
24、vocabularies. There are several factors that have contributed to this rapid progress. First, there is the coming of age of the HMM. HMM is powerfill in that, with the availability of training data, the parameters of the model can be trained automatically to give optimal performance.Second, much effo
25、rt has gone into the development of large speech corpora for system development, training, and testing. Some of these corpora are designed for acoustic phonetic research, while others are highly task specific. Nowadays, it is not uncommon to have tens of thousands of sentences available for system t
26、raining and testing. These corpora permit researchers to quantify the acoustic cues important for phonetic contrasts and to determine parameters of the recognizers in a statistically meaningful way. While many of these corpora (e.g., TIMIT, RM, APIS, and WSJ; see section 12.3) were originally collec
27、ted under the sponsorship of the U.S. Defense Advanced Research Projects Agency (ARPA) to spur human language technology development among its contractors, they have nevertheless gained world-wide acceptance (eg, in Canada, France, Germany, Japan, and the U.K.) as standards on which to evaluate spee
28、ch recognition.Third, progress has been brought about by the establishment of standards for performance evaluation. Only a decade ago, researchers trained and tested their systems using locally collected data, and had not been very careful in delineating training and testing sets. As a result, it wa
29、s very difficult to compare performance across systems, and a systems performance typically degraded when it was presented with previously unseen data. The recent availability of a large body of data in the public domain, coupled with the specification of evaluation standards, has resulted in unifor
30、m documentation of test results, thus contributing to greater reliability in monitoring progress (corpus development activities and evaluation methodologies are summarized in chapters 12 and 13 respectively).Finally, advances in computer technology have also indirectly influenced our progress. The a
31、vailability of fast computers with inexpensive mass storage capabilities has enabled researchers to run many large scale experiments in a short amount of time. This means that the elapsed time between an idea and its implementation and evaluation is greatly reduced. In fact, speech recognition syste
32、ms with reasonable performance can now run in real time using high-end workstations without additional hardwarea feat unimaginable only a few years ago.One of the most popular, and potentially most useful tasks with low perplexity (PP=11) is the recognition of digits. For American English, speaker-i
33、ndependent recognition of digit strings spoken continuously and restricted to telephone bandwidth can achieve an error rate of 0.3% when the string length is known.One of the best known moderate-perplexity tasks is the 1,000-word so-called Resource 5 Management (RM) task, in which inquiries can be made concerning various naval vessels in the Pacific
