Search

This paper presents an interactive method for designing robot facial expressions and motions of an intelligent transformable phone robot BaBi. Designing an objective function to evaluate facial expressions and motions suitable for BaBi with an original structure is quite hard although users can easily imagine favorite motions. Therefore, the proposed method employs Interactive Evolutionary Computation (IEC) in addition to general optimization. The method allows users to add candidates to a case base when finding satisfying candidates and evaluates candidates’ scores according to the case base. It applies a Different Evolution (DE) method to generate new candidates. Users may alternately push the buttons to generate face and body motion without making any selection of candidates with a selection interface. The method will make selection with the scores when users do not make any selection of candidates. Users can select or not select. It is an IEC process when users make selections, while it is a non-Interactive Evolutionary Computation (non-IEC) process when users do not make any selection, resulting in being a fusion of IEC and non-IEC. Adding more candidates to the case base can speed up generating the desired motion and promote the non-IEC progress and reduce the burden on users. Experimental results have shown that the proposed method produces the robot motions effectively.

We present an intelligent transformable phone robot named BaBi with multimodal interactions. BaBi begins with a smart phone, transforms to a movable robot from the smart phone when calling, “open”. It realizes various express to extend current smart phone to be far-field, context-driven and multimodal interactions. The new robot form that can establish both autonomous mobility and portability was proposed. Implementation of the proposed transformable robot was shown. We developed a semiconductor chip that controls wheels and servomotors of the proposed robot shape to reduce the robot footprint about the same as a smartphone. Preliminary questionnaire survey was conducted to investigate the effectiveness of the proposed robot form. When we open it (the smart phone box), it will provide us more services and more convenient functions to bring us humanized use. BaBi likes a treasure box, when opening it we can obtain a lot of goods. Therefore, we call it moonlight box that is in Chinese because we consider BaBi is a treasure. BaBi is a movable, transformable, portable, intelligent partner.

The paper presents a recognition system of online overlaid handwritten Japanese text patterns on a smart phone or baby-face tablet. The proposed system oversegments a sequence of strokes into primitive segments at candidate off-strokes between strokes using a SVM model. One or more consecutive primitive segments form a candidate character pattern, which is recognized into a list of candidate categories. Then, a segmentation and recognition candidate lattice is constructed to represent all candidate character patterns and their corresponding character classes. Finally, the optimal path is effectively found by the Viterbi search in the lattice, combining the scores of character recognition, geometric features, linguistic context, as well as the segmentation scores by SVM classification. This system incorporates feature reduction and non-character pruning to decrease the time cost per character, and semi-incremental recognition to decrease waiting time. The recognition rates on generated and collected overlaid handwritten text are 92.16% and 93.04%, respectively. The average time cost per character is not more than 0.6 s, and the average waiting time is less than 0.875 s even on an Intel Atom 1.33 GHz CPU: a low power consumption CPU for small tablets and embedded devices. Therefore, we confirm that our system recognizes online overlaid handwritten text composed of thousands of Japanese character classes with the high recognition rate without excessive waiting time.

The approximate process range for preparing the Cu&ndash, Cr&ndash, Zr alloy by selective laser melting (SLM) was determined by ANSYS simulation, and the influence of the SLM process parameters on the comprehensive properties of the SLM-formed alloy was studied by the design of experiments. The Cu&ndash, Zr alloy with optimum strength and hardness was prepared with high efficiency by optimizing the process parameters for SLM (i.e., laser power, scanning speed, and hatching distance). It is experimentally shown that tensile strength and hardness of the SLM alloy are increased by increasing laser power and decreasing scanning speed, whereas they are initially increased and then decreased by increasing the hatching distance. Moreover, strength, roughness and hardness of the SLM alloy are optimized when laser power is 460 W, scanning speed is 700 mm/s and hatching distance is 0.06 mm. The optimized properties of the SLM alloy are a tensile strength of 153.5 MPa, hardness of 119 HV, roughness of 31.384 &mu, m and relative density of 91.62%.

This paper presents an intelligent meeting recording system for an intelligent personal robot named BoBi secretary. Participants register their voices to the system before meeting. In meeting, a Microphone records voice while applying another process to recognize the voice at the same time to transform it into text resulting in an online meeting recording. The speech of the meeting recording system includes voices of multiple speakers, and it is necessary to divide it into parts of speakers before recognizing them and applying speech recognition. We attempt two methods to recognize the meeting voice: segmentation-based method and segmentation-free method. The segmentation-based method applies a Voice Activity Detection (VAD) algorithm to detect the speech and the non-speech frames to segment the recorded voice into speaker parts. Then it applies Gaussian Mixture Models (GMMs) to Mel-Frequency Cepstral Coefficient (MFCC) features extracted from each part to recognize which speaker it is from. The segmentation-free method evaluates the voice frames using MFCC features and GMMs, and then searches the optimal path as the segmentation result into the voice frames lattice. Finally each speaker's voice is sent to a speech recognition server to obtain a result text. We make a comparison between meeting recording system using segmentation-free method and that using segmentation-based method, and show the result that the segmentation-based method brings better result.

In this paper, a systematic method is described that constructs an efficient and a robust coarse classifier from a large number of basic recognizers obtained by different parameters of feature extraction, different discriminant methods or functions, etc. The architecture of the coarse classification is a sequential cascade of basic recognizers that reduces the candidates after each basic recognizer. A genetic algorithm determines the best cascade with the best speed and highest performance. The method was applied for on-line handwritten Chinese and Japanese character recognitions. We produced hundreds of basic recognizers with different classification costs and different classification accuracies by changing parameters of feature extraction and discriminant functions. From these basic recognizers, we obtained a rather simple two-stage cascade, resulting in the whole recognition time being reduced largely while maintaining classification and recognition rates. We propose a method for coarse classifier construction.It constructs a coarse classifier from 243 basic recognizers.The basic recognizers are obtained by different parameters.The architecture of the coarse classifier is a sequential cascade of basic recognizers.A genetic algorithm determines the best cascade.

This paper presents a candidate lattice refinement method for online handwritten Japanese text recognition. In the integrated segmentation-recognition framework, we first over-segment a character string pattern into primitive segments at least at their true boundaries so that each primitive segment may compose a single character or a part of a character. Then a candidate lattice is constructed based on the primitive segments. We search within the candidate lattice to obtain the optimal path as recognition result. In striving for high recognition accuracy, however, the approach must generate many candidate lattice nodes, which ultimately increase the recognition time. To solve this problem, we refine the candidate lattice to eliminate unnecessary nodes before path search and text recognition. For the refinement, we evaluate all segmentation hypotheses by combining the probability of a character verifier using noncharacter samples, the class-independent unary and binary geometric context, as well as character segmentation. We retain N-best paths by beam search to reduce the complexity of the candidate lattice. Experiments on horizontal text lines extracted from the Kondate database show that the proposed method keeps recognition accuracy while reducing recognition time to half.

This paper describes a comparison between online handwritten cursive word recognition using segmentation-free method and that using segmentation-based method. To search the optimal segmentation and recognition path as the recognition result, we attempt two methods: segmentation-free and segmentation-based, where we expand the search space using a character-synchronous beam search strategy. The probable search paths are evaluated by integrating character recognition scores with geometric characteristics of the character patterns in a Conditional Random Field (CRF) model. Our methods restrict the search paths from the trie lexicon of words and preceding paths during path search. We show this comparison on a publicly available dataset (lAM-OnDB).

The paper presents a recognition method of character-position-free (CPF) on-line handwritten Japanese text patterns to allow a user to overlay characters freely without confirming previously written characters. To develop this method, we prepared large sets of CPF handwritten Japanese text patterns artificially from normally handwritten text patterns. The proposed method sets each off-stroke between real strokes as undecided and evaluates the segmentation probability by SVM model. Then, the optimal segmentation-recognition path can be effectively found by the Viterbi search in the candidate lattice, combining the scores of character recognition, geometric features, linguistic context, as well as the segmentation scores by SVM classification. We test this method on variously overlaid sample patterns, and verify that it produces competing recognition rates as the latest recognizer for normally handwritten horizontal Japanese text without the serious problem in speed for practical applications.

This paper describes a robust context integration model for on-line handwritten Japanese text recognition. Based on string class probability approximation, the proposed method evaluates the likelihood of candidate segmentation–recognition paths by combining the scores of character recognition, unary and binary geometric features, as well as linguistic context. The path evaluation criterion can flexibly combine the scores of various contexts and is insensitive to the variability in path length, and so, the optimal segmentation path with its string class can be effectively found by Viterbi search. Moreover, the model parameters are estimated by the genetic algorithm so as to optimize the holistic string recognition performance. In experiments on horizontal text lines extracted from the TUAT Kondate database, the proposed method achieves the segmentation rate of 0.9934 that corresponds to a f-measure and the character recognition rate of 92.80%.

This paper describes an on-line handwritten Japanese text recognition system that is liberated from constraints on line direction and character orientation. The recognition system first separates freely written text into text line elements, second estimates the line direction and character orientation using the time sequence information of pen-tip coordinates, third hypothetically segment it into characters using geometric features and apply character recognition. The final step is to select the most plausible interpretation by evaluating the likelihood composed of character segmentation, character recognition, character pattern structure and context. The method can cope with a mixture of vertical, horizontal and skewed text lines with arbitrary character orientations. It is expected useful for tablet PC's, interactive electronic whiteboards and so on.

This paper presents a model and its effect for on-line handwritten Japanese text recognition free from line-direction constraint and writing format constraint such as character writing boxes or ruled lines. The model evaluates the likelihood composed of character segmentation, character recognition, character pattern structure and context. The likelihood of character pattern structure considers the plausible height, width and inner gaps within a character pattern that appear in Chinese characters composed of multiple radicals (subpatterns). The recognition system incorporating this model separates freely written text into text line elements, estimates the average character size of each element, hypothetically segments it into characters using geometric features, applies character recognition to segmented patterns and employs the model to search the text interpretation that maximizes likelihood as Japanese text. We show the effectiveness of the model through recognition experiments and clarify how the newly modeled factors in the likelihood affect the overall recognition rate.

This paper presents a semi-incremental recognition method for online handwritten English text. We employ local processing strategy and focus on a recent sequence of strokes defined as "scope". For the latest scope, we build and update a segmentation and recognition candidate lattice and advance the best-path search incrementally. We utilize the result of the best-path search in the previous scope to exclude unnecessary segmentation candidates. This reduces the number of candidate word recognition with the result of reduced processing time. We also reuse the segmentation and recognition candidate lattice in the previous scope for the latest scope. Moreover, triggering recognition processes every few strokes save CPU time. Experiment made on IAM-OnDB database shows the effectiveness of the proposed method not only in reduced processing time and waiting time, but also in recognition accuracy.

This paper presents a semi-incremental recognition method for online Japanese handwritten text recognition, which is used for busy recognition interface (recognition while writing) and lazy recognition interface (recognition after writing) without large waiting time. We employ local processing strategy and focus on a recent sequence of strokes defined as "scope". For the latest scope, we build and update a segmentation and recognition candidate lattice and advance the best-path search incrementally. We utilize the result of the best-path search in the previous scope to exclude unnecessary segmentation candidates. This reduces the number of candidate character recognition with the result of reduced processing time. We also reuse the segmentation and recognition candidate lattice in the previous scope for the latest scope. Moreover, triggering recognition processes every few strokes save CPU time. Experiment made on TUAT-Kondate database shows the effectiveness of the proposed method not only in reduced processing time and waiting time, but also in recognition accuracy.

This paper describes an online handwritten English cursive word recognition method using a segmentation-free Markov random field (MRF) model in combination with an offline recognition method which uses pseudo 2D bi-moment normalization (P2DBMN) and modified quadratic discriminant function (MQDF). It extracts feature points along the pen-tip trace from pen-down to pen-up and uses the feature point coordinates as unary features and the differences in coordinates between the neighboring feature points as binary features. Each character is modeled as a MRF and word MRFs are constructed by concatenating character MRFs according to a trie lexicon of words during recognition. Our method expands the search space using a character-synchronous beam search strategy to search the segmentation and recognition paths. This method restricts the search paths from the trie lexicon of words and preceding paths, as well as the lengths of feature points during path search. We also combine it with a P2DBMN-MQDF recognizer that is widely used for Chinese and Japanese character recognition.

Handwritten character pattern recognition methods are generally divided into two types: online recognition and offline recognition [1]. Online recognition recognizes character patterns captured from a pen-based or touch-based input device where trajectories of pentip or finger-tip movements are recorded, while offline recognition recognizes character patterns captured from a scanner or a camera device as two dimensional images.

This paper describes a method for building a compact on-line Markov random field (MRF) recognizer for large handwritten Japanese character set using structured dictionary representation and vector quantization (VQ) technique. The method splits character patterns into radicals, whose models by MRF are shared by different characters such that a character model is constructed from the constituent radical models. Many distinct radicals are shared by many characters with the result that the storage space of model dictionary can be saved. Moreover, in order to further compress the parameters, we employ VQ technique to cluster parameter sets of the mean vectors and covariance matrixes for MRF unary features and binary features as well as the transition probabilities of each state into groups. By sharing a common parameter set for each group, the dictionary of the MRF recognizer can be greatly compressed without recognition accuracy loss.

In this paper, we present methods of segmenting Nom historical documents and clustering character patterns to build a Nom character pattern database. Nom is an ideographic script to represent Vietnamese, used from the 10th century to 20th century. However, this heritage is nearly lost. In order to preserve the wisdom and knowledge expressed in Nom, recognition and digitalization are indispensable. Because there is no OCR for Nom yet, we have to start from collecting patterns. We have employed a projection profile based method for segmenting hundreds of pages into individual characters. Then, we have implemented a combination of Chinese OCR-based clustering and K-means clustering to group characters into categories. The experiment shows that the proposed system can help collecting the characters patterns effectively. Moreover, it has revealed that there are many character classes lost or uncategorized so far.

This paper presents the effect of text/non-text classification for ink search which employs string recognition. Pen or touch interfaces provides the benefit that users can write text and draw figures without changing the device or mode, but line drawings are troublesome for ink search. We propose the insertion of text/non-text classification before ink search and show its effect. For ink search, we employ our own engine to search keywords in the candidate lattice prepared by on-line handwritten Japanese text recognition, since this method produces higher search rate for Japanese text in digital ink than word spotting without ink recognition.

This paper describes a Markov random field (MRF) model with weighting parameters optimized by conditional random field (CRF) for on-line recognition of handwritten Japanese characters. It also presents updated evaluation using a large testing set. The model extracts feature points along the pen-tip trace from pen-down to pen-up and sets each feature point from an input pattern as a site and each state from a character class as a label. It employs the coordinates of feature points as unary features and the differences in coordinates between the neighboring feature points as binary features. The weighting parameters are estimated by CRF or the minimum classification error (MCE) method. In experiments using the TUAT Kuchibue database, the method achieved a character recognition rate of 92.77%, which is higher than the previous model's rate, and the method of estimating the weighting parameters using CRF was more accurate than using MCE.

In offline handwritten character recognition, the nonlinear normalization (NLN) method based on line density equalization has been proven very effective. This paper shows the effects on online handwritten Japanese character recognition. We apply the nonlinear normalization based on line density equalization to online character patterns. Since the curve-fitting-based normalization methods and their pseudo 2D extensions yields superior performance on offline patterns, we also combine these methods with the way using line density projection. We have compared the methods using trajectory-based projection with ones using line density projection. As a result, line density-based methods yield superior accuracy and a competitive time-complexity.

This paper describes an unconstrained on-line handwritten Japanese text retrieval system from character recognition candidates. The system is based on a discriminative model which integrates the scores of character recognition, segmentation and geometric context in search and retrieval, and the parameters are trained by supervised learning. Experiments on TUAT Kuchibue database show that the proposed method can effectively improve the system performance. When the search method with the optimal threshold retrieves for a keyword consisting of two, three or four characters, its f-measure is 0.720, 0.868 or 0.923, respectively.

This paper describes effective object function design for combining on-line and off-line character recognizers for on-line handwritten Japanese text recognition. We combine on-line and off-line recognizers using a linear or nonlinear function with weighting parameters optimized by the MCE criterion. We apply a k-means method to cluster the parameters of all character categories into groups so that the categories belonging to the same group have the same weight parameters. Moreover, we apply a genetic algorithm to estimate super parameters such as the number of clusters, initial learning rate and maximum learning times as well as the sigmoid function parameter for MCE optimization. Experimental results on horizontal text lines extracted from the TUAT Kondate database demonstrate the superiority of our method.

In this paper, we describe a method of retrieving on-line handwritten text based on directional feature matching. Although text search into the character recognition candidate lattice has been elaborated, the character recognition based approach does not support languages which are not assumed. The proposed method is liberated from this constraint. It first hypothetically segments on-line handwritten text into character pattern blocks and prepares the object text patterns by combining the character pattern blocks. On the other hand, it employs handwritten text as a query pattern or prepares a query pattern by combining character ink patterns from query character codes. Then, it extracts directional features from the object text patterns and the query pattern, and the dimensionalities of those features are further reduced by Fisher linear discriminate analysis (FDA). Finally, the similarity is measured between the object text patterns and the query pattern by block-shift matching. This paper discusses the retrieval performance in comparison with our previous character recognition based method.

This paper describes a Markov random field (MRF) model with weighting parameters optimized by conditional random field (CRF) for on-line recognition of handwritten Japanese characters. The model extracts feature points along the pen-tip trace from pen-down to pen-up and sets each feature point from an input pattern as a site and each state from a character class as a label. It employs the coordinates of feature points as unary features and the differences in coordinates between the neighboring feature points as binary features. The weighting parameters are estimated by CRF or the minimum classification error (MCE) method. In experiments using the TUAT Kuchibue database, the method achieved a character recognition rate of 92.77%, which is higher than the previous model's rate, and the method of estimating the weighting parameters using CRF was more accurate than using MCE.

This paper presents the idea of exam script analysis and the design to provide it as well as the support for examiners to mark exams written using pen and paper devices. It also describes partial implementation going on. Pen and paper devices capture the traces of pen movement (called digital ink), allowing the answers to be reorganized independent from paper format, analyzed and marked semi-automatically under the examiner's supervision. Furthermore, digital ink carries time stamp for each stroke so that we can use this information to analyze the question answering process of the examinees such as revisions in the answer given for a question, the time taken in answering a question and so on. This information also allows us to consider the appropriateness of the exam design and improve it.

This paper reports optimization efforts on a layered search space method aimed at accelerating the recognition of a large prototype set. The layered search space method classifies similar prototypes into clusters. Representative prototypes, each for one of these clusters, are then selected and further classified into higher-level clusters, and so on. In finding a prototype, we first identify the highest-level clusters where the prototype may be found, then proceed to identify the most likely sub- clusters within these clusters, and so on. Finally, we match the input with the prototypes in the identified lowest level clusters. Increasing layers will decrease the number of prototypes to be matched, but the precision of candidate selection will decrease and overhead will increase. Hence there are several parameters that one needs to adjust for the method to perform optimally. Most importantly, there is an optimal number of layers that accelerates the recognition without compromising the recognition rate. We used two efficient methods to approximately identify this number. Both the methods show that having two layers achieves this optimality for the recognition of handwritten Japanese characters.

This paper presents a revised method for keyword search from Japanese handwritten digital ink. We employ Japanese string recognition and produce a candidate lattice. We search for a given keyword into the lattice so that we can search for the keyword even if constituent characters are not in the top candidates. We present some overall performance as well as consideration on search errors.

This paper presents a revised method for keyword search from handwritten digital ink in comparison with the previous system. We adopt a search method using noise reduction. Experiments on digital ink databases show that the revised method typically improves the system’s overall accuracy (f-measure) from 0.653 to 0.891.

This paper describes a method of on-line handwritten Japanese text recognition by improved path evaluation. Based on a theoretical ground, the method evaluates the likelihood of candidate segmentation paths by combining scores of character pattern size, inner gap, character recognition, single and pair character position, candidate segmentation point and linguistic context, with the weight parameters optimized by a genetic algorithm. The path score is insensitive to the number of candidate patterns and the optimal path can be found by Viterbi search. Experimental results demonstrate the superiority of the proposed method.

This paper discusses online handwriting recognition of Japanese characters, a mixture of ideographic characters (Kanji) of Chinese origin, and the phonetic characters made from them. Most Kanji character patterns are composed of multiple subpatterns, called radicals, which are shared among many (sometimes hundreds of) Kanji character patterns. This is common in Oriental languages of Chinese origin, i.e., Chinese, Korean and Japanese. It is also common that each language has thousands of characters. Given these characteristics, structured character pattern representation (SCPR) composed of subpatterns is effective in terms of the size reduction of a prototype dictionary (a set of prototype patterns) and the robustness to deformation of common subpatterns. In this paper, we show a prototype learning algorithm and HMM-based recognition for SCPR. Then, we combine the SCPR-based online recognizer with a compact offline recognizer employing quadratic discriminant functions. Moreover, we also discuss online handwritten Japanese text recognition and propose character orientation-free and line direction-free handwritten text recognition and segmentation. Finally, as applications of online handwritten Japanese text recognition, we show segmentation of mixed objects of text, formulas, tables and line-drawings, and handwritten text search.

This paper describes a method of producing segmentation point candidates for on-line handwritten Japanese text by a support vector machine (SVM) to improve text recognition. This method extracts multi-dimensional features from on-line strokes of handwritten text and applies the SVM to the extracted features to produces segmentation point candidates. We incorporate the method into the segmentation by recognition scheme based on a stochastic model which evaluates the likelihood composed of character pattern structure, character segmentation, character recognition and context to finally determine segmentation points and recognize handwritten Japanese text. This paper also shows the details of generating segmentation point candidates in order to achieve high discrimination rate by finding the combination of the segmentation threshold and the concatenation threshold. We compare the method for segmentation by the SVM with that by a neural network using the database HANDS-Kondate_t_bf-2001-11 and show the result that the method by the SVM bring about a better segmentation rate and character recognition rate.

This paper describes prototyping of a form processingsystem employing dot texture for printing input frames ofthe form. The dot texture is the texture composed of smalldots. It eases the separation of handwritings from theinput frames even under monochrome printing/readingenvironments and makes the system to process thehandwritings according to the information embedded inthe dot texture of the frames. The embedded informationin the form dictates how to process the form so that wecall the form "active form" being opposite to the passiveform processed by the program stored in a documentreader. This method can also be used to embed otherinformation such as attribute of handwriting and so on.This paper presents the design, prototyping and somepreliminary evaluation.