摘要:
In a system for recognizing a time sequence of feature vectors of a speech signal representative of an unknown utterance as one of a plurality of reference patterns, a generator (11) for generating the reference patterns has a converter (15) for converting a plurality of time sequences of feature vectors of an input pattern of a speech signal with variances to a plurality of time sequences of feature codes with reference to code vectors (14) which are previously prepared by the known clustering. A first pattern former (16) generates a state transition probability distribution and an occurrence probability distribution of feature codes for each state in a state transition network. A function generator (17) calculates parameters of continuous Gaussian density function from the code vectors and the occurrence probability distribution to produce the continuous Gaussian density function approximating the occurrence probability distribution. A second pattern former (18) produces a reference pattern defined by the state transition probability distribution and the continuous Gaussian density function. For a plurality of different training words, a plurality of reference patterns are generated and are memorized in the reference pattern generator.
摘要:
A speech recognition apparatus of the speaker adaptation type operates to recognize an inputted speech pattern produced by a particular speaker by using a reference pattern produced by a voice of a standard speaker. The speech recognition apparatus is adapted to the speech of the particular speaker by converting the reference pattern into a normalized pattern by a neural network unit, internal parameters of which are modified through a learning operation using a normalized feature vector of the training pattern produced by the voice of the particular speaker and normalized on the basis of the reference pattern, so that the neural netowrk unit provides an optimum output similar to the corresponding normalized feature vector of the training pattern. In the alternative, the speech recognition apparatus operates to recognize an inputted speech pattern by converting the inputted speech pattern into a normalized speech pattern by the neural network unit, internal parameters of which are modified through a learning operation using a feature vector of the reference pattern normalized on the basis of the training pattern, so that the neural network unit provides an optimum output similar to the corresponding normalized feature vector of the reference pattern and recognizing the normalized speech pattern according to the reference pattern.
摘要:
An on-line character recognition method is disclosed that recognizes inputted characters on-line by finding distance between strokes for patterns in stroke units of inputted characters and patterns in stroke units for each reference stroke. Reference patterns and inputted character patterns are each divided and represented as stroke shape patterns that indicate the shapes of strokes and stroke position patterns that indicate the position or size of strokes. Inter-stroke shape distances corresponding to each stroke shape pattern and inter-stroke position distances corresponding to each stroke position pattern are found, following which the inter-stroke distance is found based on the inter-stroke shape distances and the inter-stroke position distances.
摘要:
An online character recognition system comprises a standard stroke storing unit for storing a standard stroke and its correlated standard stroke number, a standard stroke number of strokes storing unit for storing number of strokes information indicating in how many stroke character each standard stroke appears, a character dictionary storing unit for storing a category to be recognized and its correlated standard stroke number-string, a standard stroke control unit for referring to the standard stroke number of strokes storing unit to selectively read, from the standard stroke storing unit, a standard stroke having the same number of strokes information as the number of strokes of an input character, an inter-stroke distance calculating unit for calculating an inter-stroke distance between a standard stroke read by said standard stroke control unit and an input stroke, and a matching unit for recognizing an input character based on an inter-stroke distance calculated by the inter-stroke distance calculating unit.
摘要:
An organism collation apparatus capable of randomly selecting type of organism information to be collated is disclosed. A plurality of types of organism information owned by one human being like fingerprint patterns of the ten fingers are registered in advance for the different types like the “thumb of the right hand”. One type is selected from among the plurality of types of registered organism information, and inputting of organism information of the selected type is indicated to a person to be collated. In this condition, inputted organism information is accepted, and the inputted organism information and the registered organism information of the selected type are collated with each other.
摘要:
To provide a speech recognition apparatus which enables the reduction of transmission time and of costs. A terminal-side apparatus (100) includes a speech detection portion (101) for detecting a speech interval of inputted data, a waveform compression portion (102) for compressing waveform data at the detected speech interval, and a waveform transmission portion (103) for producing the compressed waveform data. A server-side apparatus (200) includes a waveform reception portion (201) for receiving the waveform data transmitted from the terminal-side apparatus, a waveform decompression portion (202) for decompressing the received waveform data, an analyzing portion (203) for analyzing the decompressed waveform data, and a recognizing portion (204) for performing recognition processing to produce a recognition result.
摘要:
A conventional speech recognition network finite-state automaton, which follows regular grammar rules, is improved by adding subnetworks tapped into the original network at call and return points, whereby context-free grammar rules may be used, with avoidance of infinite loop response of a recurrent expression.A continuous speech recognition apparatus includes a standard pattern memory for storing standard patterns, a distance calculating section for calculating distances between frames of an input speech pattern and the standard patterns, an accumulation value calculating section for calculating accumulation values of the distances on matching paths which cause frames of the speech pattern and the standard patterns to correspond to each other, an accumulation value memory for storing the accumulation values, a return point memory for storing an address of a return point of a subnetwork in correspondence with the same address as that of the accumulation value memory, a call processing section for writing a minimum value of the accumulation values at a plurality of call points for the subnetwork as an initial value of the accumulation value for the subnetwork in the accumulation value memory and writing an address of a return point corresponding to the call point yielding the minimum value in the return point memory as an initial value, and a return processing section for writing an accumulation value at a terminal point of the subnetwork in the accumulation value memory addressed by the return point address stored in the return point memory at the terminal point of the subnetwork.
摘要:
A continuous speech recognition unit using forward probabilities for recognizing continuous speech associated with standard patterns for given units of recognition comprises a standard template memory for storing Markov model standard templates of standard speech, which are composed of state sequences and transition probabilities between the states; an observation probability computing device for computing a forward probability for a feature vector time sequence; and a cumulative value computing device for determining a cumulative value based on the sum of previous cumulative values. The unit further comprises a matching pass memory for storing maximum values produced by the cumulative value computing means and a result processor for determining recognition results indicative of recognized words. The unit stores the transition giving the best probability in memory for each state and traces back the recognition result for the word sequence based on the transitions in memory.
摘要:
A recognition system for recognizing a plurality of continuous hand-written characters, employing a first memory in which isolated characters are stored, and a second memory which stores information, including interstroke character information, for connecting isolated characters. According to various embodiments of the invention, this interstroke information may be stored as part of a continuous character, or by itself.