scispace - formally typeset
Search or ask a question

Showing papers on "Feature (machine learning) published in 1985"


Book ChapterDOI
01 Jan 1985
TL;DR: Only pattern matching in case expressions is described here, since the LML compiler project began as an attempt to produce efficient code for a typed functional language with lazy evaluation and it should also yield efficient code.
Abstract: Introduction Pattern matching is a very powerful and useful device in programming. In functional languages it emerged in SASL [Turn76] and Hope [BursS0], and has also found its way into SML [Miln84]. The pattern mathing described here is that of LML which is a lazy ([Frie76] and [Henri76]) variant of ML. The pattern matching in LML evolved independently of that in SML so they are not (yet) the same, although very similar. The compilation of pattern matching in SML has been addressed in [Card84]. The LML compiler project began as an attempt to produce efficient code for a typed functional language with lazy evaluation. Since we regard pattern matching as an important language feature it should also yield efficient code. Only pattern matching in case expressions is described here, since we regard this as the basic pattern matching facility in the language. All other types of pattern mathing used in LML can be easily translated into case expressions, see [Augu84] for details. The compilation (of pattern matching) proceeds in several steps: • transform all pattern matching to case expressions. • transform complex case expressions into expressions that are easy to generate code for. • generate G-code for the case expressions, and from that machine code for the target machine.

200 citations


Proceedings ArticleDOI
18 Aug 1985
TL;DR: A system is described that integrates vision and tactile sensing in a robotics environment to perform object recognition tasks that uses multiple sensor systems to compute three dimensional primitives that can be matched against a model data base of complex curved surface objects containing holes and cavities.
Abstract: A system is described that integrates vision and tactile sensing in a robotics environment to perform object recognition tasks. It uses multiple sensor systems (active touch and passive stereo vision) to compute three dimensional primitives that can be matched against a model data base of complex curved surface objects containing holes and cavities. The low level sensing elements provide local surface and feature matches which are constrained by relational criteria embedded in the models. Once a model has been invoked, a verification procedure establishes confidence measures for a correct recognition. The three dimen* sional nature of the sensed data makes the matching process more robust as does the system's ability to sense visually occluded areas with touch. The model is hierarchic in nature and allows matching at different levels to provide support or inhibition for recognition.

107 citations


Journal ArticleDOI
TL;DR: The paper reviews salient issues in the application of conventional techniques for extraction of information and the systems that use the artificial intelligence approach are characterized with respect to three key properties.

62 citations


Patent
10 Oct 1985
TL;DR: In this article, a method for identifying unknown input data, such as patterns or characters, is proposed, where a large number of reference data is collected and analyzed in order to form "ringed clusters" for each class of input data.
Abstract: A method for identifying unknown input data, such as patterns or characters. In order to classify unknown input characters, first, during a preprocessing phase, large number of reference data is collected and analyzed in order to form "ringed clusters" for each class of input data. For example, if the input data are characters, a set of ringed clusters is associated with each character class, such as all "e". These ringed clusters may be coarse, medium or fine depending upon the desired accuracy in classifying the input characters. The ringed clusters include "certainty spheres" which are used to identify with certainty an unknown input character if it lies within such a sphere. The ringed clusters also include "confidence spheres" which are used to identify, although not with certainty, the unknown input character, and assign a confidence value indicating the relative confidence associated with the possibility that this unknown character corresponds to the reference data class of the ringed cluster.

56 citations


Journal ArticleDOI
01 Mar 1985
TL;DR: It is shown that features derived from facets and Gaussian curvature are effective in the classification of solder joints using a minimum-distance classification algorithm.
Abstract: An approach is described for the automatic inspection of solder joints on printed circuit boards. Common defects are identified in solder joints and a joint is classified as being good or belonging to one of the defective classes. The motivation for this classification is not just the detection of defective joints, but the desire to automatically take corrective action on the assembly line. The features used for classification are based on characteristics of intensity surfaces. It is shown that features derived from facets and Gaussian curvature are effective in the classification of solder joints using a minimum-distance classification algorithm. Class separation plots are shown to be useful for quickly studying individual effectiveness of a feature or pair of features in classification. Results show the efficacy of the described approach.

56 citations


Patent
17 Sep 1985
TL;DR: In this article, a pattern recognition device is arranged to have learning of a reference pattern vector carried out in a recognition unit by making use of the pause periods in the recognition processing, without particularly providing a learning section for learning the reference pattern vectors.
Abstract: A pattern recognition device is arranged to have learning of a reference pattern vector carried out in a recognition unit by making use of the pause periods in the recognition processing, without particularly providing a learning section for learning the reference pattern vector. Namely, a part or the entirety of the arithmetic processing unit where the recognition result is obtained in the recognition unit by collating the input pattern with the recognition dictionary, can be utilized as the learning portion of the reference pattern vector. In concrete terms, the operation of multiplication-accumulation (inner product) which represents the main operation in the recognition processing and the learning processing, can be carried out by means of similar processes of handling. Therefore, by utilizing the processing section for sum of the products operation, of the recognition unit, which is in the idle state for pattern recognition, it becomes possible to carry out learning the reference pattern vector efficiently in time, without forcibly interrupting the pattern recognition processing by formally setting a learning condition.

56 citations


Journal ArticleDOI
01 Jan 1985
TL;DR: A new set of topological features (primitives) for use with a syntactic classifier for high-accuracy recognition of handwritten numerals is proposed and made possible with minimal preprocessing of the test pattern.
Abstract: A new set of topological features (primitives) for use with a syntactic classifier for high-accuracy recognition of handwritten numerals is proposed. The tree grammar used in this study makes it possible to achieve high-recognition speeds with minimal preprocessing of the test pattern.

47 citations


Patent
27 Jul 1985
TL;DR: In this paper, a comparison operation circuit uses a new picture element and a dictionary vector being the result of preceding operation in a storage device 4 for the purpose of operation every time a new specific element of a feature parameter Pi is given from a feature extraction circuit 1 with the operation of a control circuit 9 and the dictionary vector is stored in storage device 5.
Abstract: PURPOSE:To attain ease of forming of a dictionary by comparing and operating plural extracted feature parameters and a dictionary, storing the result and using a dictionary vector selected from the storage as the result of recognition so as to realize an excellent pattern recognition. CONSTITUTION:A comparison operation circuit 3 uses a new picture element and a dictionary vector being the result of preceding operation in a storage device 4 for the purpose of operation every time a new specific element of a feature parameter Pi is given from a feature extraction circuit 1 with the operation of a control circuit 9 and the dictionary vector is stored in a storage device 5. Then the dictionary vector stored in the storage devices 4, 5 is inputted to a maximum similarity decision circuit 8 and when the selected dictionary is one, its dictionary vector is outputted as the result of recognition. Moreover, when plural dictionary vectors are obtained, a vector having the maximum similarity is selected and used as the result of recognition.

47 citations


Journal ArticleDOI
TL;DR: A distributed rule-based system for automatic speech recognition is described and experiments on the automatic segmentation and recognition of phrases, made of connected letters and digits, are described and discussed.
Abstract: A distributed rule-based system for automatic speech recognition is described. Acoustic property extraction and feature hypothesization are performed by the application of sequences of operators. These sequences, called plans, are executed by cooperative expert programs. Experimental results on the automatic segmentation and recognition of phrases, made of connected letters and digits, are described and discussed.

47 citations


Journal ArticleDOI
Daniel Sabbah1
TL;DR: This paper summarizes the initial foray in tackling Artificial Intelligence problems using a connectionist approach to represent and recognize projected line drawings of Origami objects and what advantages such an approach would have.

45 citations


Journal ArticleDOI
TL;DR: The correlation approaches emphasize the use of synthetic discriminant functions to achieve multiclass distortion-invariant pattern recognition and feature-extraction and correlation architectures are chosen as the two major approaches.
Abstract: Recent progress in coherent optical pattern recognition is reviewed. Emphasis is given to techniques for multiclass distortion-invariant pattern recognition. Real-time and practical architectures and algorithms are described, and recent results on extensive data bases are noted. Feature-extraction and correlation architectures are chosen as the two major approaches. The feature-extraction techniques considered include Fourier coefficients, chord distributions, and moments. The correlator approaches emphasize the use of synthetic discriminant functions to achieve multiclass distortion-invariant pattern recognition.

PatentDOI
TL;DR: It becomes possible to prepare highly reliable reference pattern vectors in an easy manner from a small number of speech patterns, which makes it possible to achieve an improvement in the speech recognition factor.
Abstract: The learning method of reference pattern vectors for speech recognition in accordance with the present invention, a plurality of speech feature vectors are generated from the time series of speech feature parameter for the input speech pattern, by taking into account knowledge concerning the variation tendencies of the speech patterns, and the learning (preparation) of reference pattern vectors for speech recognition is carried out by the use of these speech feature vectors thus generated. Therefore, it becomes possible to prepare highly reliable reference pattern vectors in an easy manner from a small number of speech patterns, which makes it possible to achieve an improvement in the speech recognition factor. In particular, it becomes possible to plan an easy improvement of the reference pattern vectors by an effective use of a relatively small number of input speech patterns.

Patent
28 Mar 1985
TL;DR: An apparatus for recognizing and displaying handwritten characters and figures in which input stroke information on a handwritten input character or figure is read out by an electromagnetic tablet, recognition means performs character/figure recognition on the basis of the feature of the input strokes as mentioned in this paper.
Abstract: An apparatus for recognizing and displaying handwritten characters and figures in which input stroke information on a handwritten input character or figure is read out by an electromagnetic tablet, recognition means performs character/figure recognition on the basis of the feature of the input stroke information, display means displays the input stroke information and the result of recognition, and when the result of recognition is displayed on a display screen of the display means, stroke information having been used for recognition is erased from the display screen, and stroke information which is not yet used for recognition, is left on the display screen.

Journal ArticleDOI
Fred W. M. Stentiford1
TL;DR: In this article, an automatic evolutionary search is applied to the problem of feature extraction in an OCR application and a performance measure based on feature independence is used to generate features which do not appear to suffer from peaking effects.
Abstract: An automatic evolutionary search is applied to the problem of feature extraction in an OCR application. A performance measure based on feature independence is used to generate features which do not appear to suffer from peaking effects [17]. Features are extracted from a training set of 30 600 machine printed 34 class alphanumeric characters derived from British mail. Classification results on the training set and a test set of 10 200 characters are reported for an increasing number of features. A 1.01 percent forced decision error rate is obtained on the test data using 316 features. The hardware implementation should be cheap and fast to operate. The performance compares favorably with current low cost OCR page readers.

Patent
23 Jul 1985
TL;DR: In this paper, the deviation pattern is defined as the deviation from the reference of repeated utterances of the reference speakers or the specified speaker, which is a measure of the similarity of the input and reference patterns, according to one of several possible distance formulae.
Abstract: A pattern recognition apparatus for recognizing spoken words of a nonspecific speaker or of a specific speaker. A reference pattern composed of a sequence of feature vectors, each composed of n feature parameters, bi, is stored. The reference pattern represents a form of average of said words to be recognized as determined by multiple reference speakers speaking the same words or by the specified speaker speaking said words several times. A deviation pattern composed of a sequence of feature vectors composed of n feature parameters, wi /2, is stored. The deviation pattern is a measure of the deviation from the reference of the repeated utterances of the reference speakers or the specified speaker. An input pattern, representing the utterances of a speaker, is composed of a sequence of feature vectors, each composed of n feature parameters, ai, and is stored. A measure of the similarity of the input and reference patterns is calculated, taking into account the deviation pattern, according to one of several possible distance formulae. Basically, a distance parameter calculated for each corresponding input, reference and deviation parameter is set to zero value if the input parameter is inside the deviation range of the reference parameter, and is otherwise calculated to be a finite value.

Proceedings ArticleDOI
01 Apr 1985
TL;DR: The speech recognition accuracy of this method in recognizing non-training voice data was 95.8% with automatic segmentation, and the category of the nearest reference pattern is taken as the result.
Abstract: This paper descrives recognition method, reference pattern generation method, and evaluation about the speaker independent recognition for telephone speech response systems. Input utterance is analyzed by 19 channel BPFs. The power and vocal cord source characteristics are normalized. The time normalization is realized by linearly compressing or expanding to 32 frames. The speech pattern undergoes pattern matching with male and female reference patterns, and the category of the nearest reference pattern is taken as the result. It is necessary to optimize the reference patterns so that the speech can be correctly recognized in spite of the difference of formant frequencies, and slight segmentation errors. To optimize the reference patterns, the recognition of the training patterns and updating of the reference patterns are repeated. A total of 256 male and female reference patterns were generated The speech recognition accuracy of this method in recognizing non-training voice data was 95.8% with automatic segmentation.

Journal ArticleDOI
TL;DR: It has been proved that the accuracy of the results by fuzzy pattern recognition is satisfactory and the method of calculation has been simplified, so that it will be advantageous to the realization of real time recognition and self-adaptive control of the metal cutting states in the future.

Proceedings ArticleDOI
01 Apr 1985
TL;DR: This paper describes the speaker independent large vocabulary speech recognition system based on phoneme recognition, which employs LPC cepstrum coefficients as the feature parameter and statistical distance measure between an input pattern and phoneme reference template.
Abstract: This paper describes the speaker independent large vocabulary speech recognition system based on phoneme recognition. Phoneme recognition employs LPC cepstrum coefficients as the feature parameter and statistical distance measure between an input pattern and phoneme reference template. Using power dips of low and high frequency range, similarity to unvoiced feature and similarity to nasal feature, the consonant segments are detected. The discrimination of phonemes is performed individually for vowels, semi-vowels and consonants. Phoneme sequence which is result of phoneme recognition is matched with each item of the word dictionary and the item with the highest similarity in the dictionary is output as the recognition result. An average phoneme recognition score is 81.4% for 212 words uttered by forty speakers including males and females; 90.6% for vowels, 78.0% for semivowels and 71.9% for consonants. An average score of word recognition is 95.6% for 274 Japanese city names uttered by forty speakers.

Journal ArticleDOI
TL;DR: This integrated circuit has been designed to be suitable for both isolated word recognition and connected speech recognition and has been tested in a pilot isolated-word and connected-speech recognition system.
Abstract: Many speech recognition systems contain the foUowing functional blocks: a voice input circuit, a feature extractor (analyzer), a unit for calculating the distance between input and standard patterns at every frame, a memory for storing standard patterns, a unit for matching whole word patterns (a pattern matching circuit), and a final decision and system control circuit. The pattern matching circuit is independent of the recognition algorithm and requires many conventional integrated circuits for implementation. Therefore, we decided to develop a custom integrated circuit for the pattern matching function. Due to its original architecture, our integrated circuit is able to execute in real time a continuous nonlinear matching process and is able to handle large volumes of data. This integrated circuit has been designed to be suitable for both isolated word recognition and connected speech recognition. The circuit has been tested in a pilot isolated-word and connected-speech recognition system. In this paper, the authors describe this integrated circuit's specifications, architecture, and performance, as well as its application in the model system.

Proceedings ArticleDOI
19 Dec 1985
TL;DR: A speech recognition time warping algorithm is adapted to picture analysis to recognize patterns despite variations in scale and orientation so that objects may be recognized regardless of whether they are embedded in other parts or they are distorted.
Abstract: The aim of this study is to adapt a speech recognition time warping algorithm to picture analysis. Our goal is to recognize patterns despite variations in scale and orientation. We may recognize objects regardless of whether they are embedded in other parts or they are distorted. The programs input real pictures, extract the contours and then encode and compare them to a pattern dictionary. The computer time is particularly short for such a recognition process.

Patent
18 Oct 1985
TL;DR: In this paper, the authors proposed a method to reduce the standard character pattern capacity, and to recognize characters with a high precision even if the object is hand-written KANJI (Chinese character), by extracting directional features where a measured character line width is used to normalize the line width.
Abstract: PURPOSE: To reduce the standard character pattern capacity, and to recognize characters with a high precision even if the object is hand-written KANJI (Chinese character), by extracting directional features where a measured character line width is used to normalize the line width. CONSTITUTION: A line width measuring part 4 measures an average line width in accordance with the number of black dots of an input character pattern. A directional feature face generating part 5 obtains a directional feature pattern in accordance with directional components of outline points of the input character pattern and uses the measured line width to normalize the line width of the directional feature pattern. Further, the blurring processing is performed by a blurring processing part 8 to extract a feature pattern. Thus, feature points such as end points, branch points, and bending points are not expressed on a display and the feature pattern where the variance of the character line width is absorbed is extracted, and therefore, a high recognition precision is obtained without increasing the number of standard patterns even if the recognition object is hand-written KANJI where feature points and the line width are remarkably varied. COPYRIGHT: (C)1988,JPO&Japio

Book ChapterDOI
10 Jul 1985
TL;DR: In this article, a machine induction program (WITT) is proposed to model human categorization. But it is not an alternative to traditional Artificial Intelligence (AI) approaches to generalization and conceptual clustering which tend to focus on necessary and sufficient feature rules.
Abstract: This paper reports a machine induction program (WITT) which attempts to model human categorization. Properties of categories that human subjects are sensitive to include, best or prototypical members, relative contrasts between putative categories, and polymorphy (niether necessary or sufficient features). This approach represents an alternative to traditional Artificial Intelligence (AI) approaches to generalization and conceptual clustering which tend to focus on necessary and sufficient feature rules, equivalence classes, and search and match algorithms. The present approach is shown to be more consistent with human categorization while potentially including results produced by more traditional clustering schemes. Applications of this categorization approach are also discussed in the domains of Expert systems and Information retrieval.

Patent
13 Aug 1985
TL;DR: In this article, a character segmenting part 13 segments a row pattern including characters for one line by the pattern memory 12, and while moving a remark point in the row direction, executes the scanning of the column direction and takes out data (black string data) obtained by indicating a part including the pattern by the number of black picture elements.
Abstract: PURPOSE:To attain accurate character reading even if characters having difference size are contacted each other by determining the number of divisions in accordance with the size when the lump of a black string on a character line is larger than a fixed section, and executing different charactor segmenting methods each other. CONSTITUTION:Characters on a form are converted into binary pattern data by a photoelectric conversion circuit and temporarily stored in a pattern memory 12 from an input terminal 11 to a pattern memory 12. A character segmenting part 13 segments a row pattern including characters for one line by the pattern memory 12, and while moving a remark point in the row direction, executes the scanning of the column direction and takes out data (black string data) obtained by indicating a part including the pattern by the number of black picture elements. In addition, the character segmenting part 13 segments an individual pattern or a forced separating pattern as a discrimination pattern from the row pattern on the basis of the black string data and sends the segmented pattern to a feature extracting part 14. The feature extracting part 14 extracts the feature of the character, a discrimination part 15 collates the extracted feature with a discrimination dictionary part 16 and a character decision part 17 processes the sent data and outputs the selected character as a character reading result.

Journal ArticleDOI
TL;DR: A method to generate synthetic training data is described, which alleviates the problem of insufficient training data and a means is provided for injecting a priori geologic knowledge into the classifier, including well logs.

Journal ArticleDOI
TL;DR: In this article, syntactic pattern recognition techniques are applied to the analysis of one-dimensional seismic traces for classification of Ricker wavelets, and the relation between error probability and Levenshtein distance is proposed.
Abstract: Syntactic pattern recognition techniques are applied to the analysis of one‐dimensional seismic traces for classification of Ricker wavelets. The system for one‐dimensional seismic analysis includes a likelihood ratio test, optimal amplitude‐dependent encoding, probability of detecting the signal involved in the global and local detection, plus minimum‐distance and nearest‐neighbor classification rules. The relation between error probability and Levenshtein distance is proposed.

Journal ArticleDOI
TL;DR: The claim that overselectivity in feature processing underlies the disorders that aphasics display in processing both visual and verbal material was directly tested and supported the assumption that feature-processing disability is a specific and separable deficit, although related both to naming ability and to severity of aphasia.

Journal ArticleDOI
TL;DR: This paper presents a flexible character recognition method with the emphasis on preprocessing and feature extraction, and results when a feature system are selected by man and when generated by machine are compared.
Abstract: This paper presents a flexible character recognition method with the emphasis on preprocessing and feature extraction. Spatial band-pass filtering and a white-black decision rule, are applied to extract the different stroke segments of characters. The optical system as a low-pass isotropic filter fulfills the basic operation of smoothing. The analog electrical system operates with the video signals on the output of the integrated photosensing array. A structural-statistical algorithm is used for character recognition. Feature set is generated automatically by means of character segmentation into structural parts and clustering. Filter modelling computer simulation results and simulation results of character recognition are presented. Character recognition results when a feature system are selected by man and when generated by machine are compared.

Journal ArticleDOI
TL;DR: A pattern recognition algorithm is described that learns a transition net grammar from positive examples that makes a step in that direction by providing a robust parser and reducing special interaction for introduction of new words and terms.
Abstract: A pattern recognition algorithm is described that learns a transition net grammar from positive examples. Two sets of examples -- one in English and one in Chinese -- are presented. It is hoped that language learning will reduce the knowledge acquisition effort for expert systems and make the natural language interface to database systems more transportable. The algorithm presented makes a step in that direction by providing a robust parser and reducing special interaction for introduction of new words and terms.

Patent
09 Apr 1985
TL;DR: In this paper, the authors proposed a method to realize character recognition which shortens a decision time by performing decision processing by using a dictionary for classification, and selecting a feature in a pair with the dictionary for detail discrimination to discriminate a similar pattern on the basis of only the feature value.
Abstract: PURPOSE:To realize character recognition which shortens a decision time by performing decision processing by using a dictionary for classification which consists of the hierachy of the same features as the features in a dictionary for detail discrimination, and selecting a feature in a pair with the dictionary for detail discrimination to discriminate a similar pattern on the basis of only the feature value. CONSTITUTION:When an unknown character is inputted, a feature extraction part 1 obtains its feature value and a classification part 2 performs a classification processing by using the dictionary for the classification. A detail discrimination part 3 calculates the similarity by using the detail discrimination dictionary at a detail discrimination dictionary part 5 and the detail discrimination dictionary limited by the classification part 2, and output the decision result based upon the similarity. A tournament decision part 4 performs a tournament processing on the basis of the decision result of the detail discrimination part 3 and outputs a final decision result. Consequently, the decision processing time is shortened by the dictionary for the classification consisting of the hierachy of the same features with the dictionary for detail discrimination and an effective feature is selected in a pair with the dictionary for detail discrimination to make a decision on the basis of only the selected feature value, discriminating a similar pattern.

Patent
27 Sep 1985
TL;DR: In this article, the authors presented a method to obtain easily a standard pattern having high recognition performance by executing the learning of the standard pattern by a KL expansion which has used a power method.
Abstract: PURPOSE: To obtain easily a standard pattern having high recognition performance by executing the learning of the standard pattern by a KL expansion which has used a power method. CONSTITUTION: A vowel pattern and a consonant pattern based on its feature parameter are stored in a learning pattern memory part 6 from an input voice through a feature extracting part 2, a segmentation processing part 3, etc. Based on this stored contents, a learning processing part 7 provided with a microprocessor part and a hardware part generates a covariance matrix of an input voice pattern, and subsequently, executes a KL expansion which has used a power method, and determines an intrinsic value and an intrinsic vector to its covariance matrix. Next, based on the intrinsic value and the intrinsic vector, a standard pattern of a specified category of a high recognition performance, which has taken a statistical distribution into consideration is stored in a standard pattern memory part 5. In such a way, by the KL expansion which has used the power method, the intrinsic value and the intrinsic vector to the covariance matrix are outputted successively from those of a large absolute value, and from a small number of patterns, a standard pattern which has reflected a statistical variation of the pattern can be derived easily at a high speed. COPYRIGHT: (C)1987,JPO&Japio