摘要:

A filtering scheme that utilizes the phase‐frequency characteristic of a filter is shown to have certain advantages for analyzing signals, such as speech, which are produced by shock or noise excitation of a physical system having one or more resonances. The phase principle is shown to be particularly well suited to neural mechanisms of inhibition and facilitation, and evidence is presented that such a principle may be used by the ear to achieve its analysis of the sound that it receives.

ISSN:0001-4966    

DOI:10.1121/1.1906937

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

The process of speech perception is analyzed into three main operations: (1) translation of the speech signal into form suitable for the nervous system, (2) identification of discrete speech elements, and (3) comprehension of meaning. The first operation appears to correspond roughly to the transformation made by the sound spectrograph. The second may be carried out by the neural equivalent of a set of matched filters. The third appears to involve a neural form of cross‐correlation that exhibits some of the properties of the analogous electronic process.

ISSN:0001-4966    

DOI:10.1121/1.1906938

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

Shannon and others have estimated that written English is about 60 percent redundant. These estimates are arrived at by considering linguistic constraints on our use of speech symbols; they do not consider additional restrictions imposed by the audience and the situation in which the speaker finds himself. In order to estimate the effects of such nonlinguistic constraints, an informational analysis has been made of the “sublanguage” used in the control of aircraft by Air Force control tower operators. When the situational, as well as linguistic, contexts are taken into account, the estimated redundancy is raised to 96 percent.

ISSN:0001-4966    

DOI:10.1121/1.1906939

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

Synthetic methods applied to isolated syllables have permitted a systematic exploration of the acoustic cues to the perception of some of the consonant sounds. Methods, results, and working hypotheses are discussed.

ISSN:0001-4966    

DOI:10.1121/1.1906940

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

Phonemes are quite abstract units. The criteria for establishing them are primarily distributional and articulatory, not acoustic. A unit intermediate between a single speech‐event‐segment and a phoneme is the allophone. An allophone is a statistical average, with constant, identifiable acoustic characterization. For many purposes of engineering, psychology, pedagogy, and information theory, the allophone may be a more suitable unit than the phoneme.

ISSN:0001-4966    

DOI:10.1121/1.1906941

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

Phonemic analysis must be based to a large extent on considerations of phonetic similarity and difference. For this purpose various stages of the speech event may be chosen, in particular the articulatory, the acoustic, and the auditory stage. It does not seem irrelevant which stage is chosen, for the analysis may give different results in each of the three stages. However, it is usually not possible to state general reasons for preferring one stage over another. In a given language functional reasons may in some cases decide the choice.

ISSN:0001-4966    

DOI:10.1121/1.1906942

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

The phonemes of a language are neither absolutes nor bundles of absolute characteristics, but rather are fluctuating bundles of features identified (1) relative to each other in sequences, (2) relative to a system of fluctuating bundles of characteristics, and (3) relative to structural position in a sequence of such relative elements. Detection techniques, if paralleling phonemic analysis, would need to be able to work with fluctuating relative elements rather than with absolute physical characteristics only.

ISSN:0001-4966    

DOI:10.1121/1.1906943

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

The development of devices which can be operated automatically from the phonetic content of speech may be viewed in terms of the more general problem of the reduction of channel capacity in communications systems. Significance has been observed in formant positions and movements as regard the identification of speech sound. The basic problems in the derivation of phonemes from the formant patterns are reviewed.

ISSN:0001-4966    

DOI:10.1121/1.1906944

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

This study deals with those aspects of speech which are phonetically significant. A technique has been developed with which phonetically equivalent speech samples may be obtained different phonetic contexts and from different speakers. Data on two front vowels by different types of speakers are presented. The technique has also been applied to the evaluation of words containing these two vowels.

ISSN:0001-4966    

DOI:10.1121/1.1906945

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP

摘要:

The recognizer discussed will automatically recognize telephone‐quality digits spoken at normal speech rates by a single individual, with an accuracy varying between 97 and 99 percent. After some preliminary analysis of the speech of any individual, the circuit can be adjusted to deliver a similar accuracy on the speech of that individual. The circuit is not, however, in its present configuration, capable of performing equally well on the speech of a series of talkers without recourse to such adjustment.Circuitry involves division of the speech spectrum into two frequency bands, one below and the other above 900 cps. Axis‐crossing counts are then individually made of both band energies to determine the frequency of the maximum syllabic rate energy with each band. Simultaneous two‐dimensional frequency portrayal is found to possess recognition significance. Standards are then determined, one for each digit of the ten‐digit series, and are built into the recognizer as a form of elemental memory. By means of a series of calculations performed automatically on the spoken input digit, a best match type comparison is made with each of the ten standard digit patterns and the digit of best match selected.

ISSN:0001-4966    

DOI:10.1121/1.1906946

出版商:Acoustical Society of America    

年代:1952

数据来源: AIP