This paper presents the results of PERSIAN HANDWRITTEN WORD RECOGNITION based on Mixture of Experts technique. In the basic form of ME the problem space is automatically divided into several subspaces for the experts, and the outputs of experts are combined by a gating network. In our proposed model, we used Mixture of Experts Multi Layered Perceptrons with Momentum term in the classification Phase. Applying this term makes three effects in our system: a) increase convergence rate, b) obtain the optimum performance in our system, c) and escape from the local minima on the error surface. We produce three different Mixture of Experts structure. Experimental result for proposed method show an error rate reduction of 6.42% compare to the mixture of MLPs experts. Comparison with some of the most related methods indicates that the proposed model yields excellent RECOGNITION rate in HANDWRITTEN WORD RECOGNITION.

In this paper a RECOGNITION system for PERSIAN WORDs is introduced which utilizes the local higher order of the log-polar image autocorrelation for feature extraction of PERSIAN sub-WORDs. This feature extraction technique brings up leads to system robustness in cases of writing variations alteration like scaled or rotated handwritings. Also using the log-polar transform, the sub-WORD image sampling will be performed so that most of acquired samples will be centered in a certain area. The proposed method uses the discrete Hidden Markov’s Model (HMM) as a classifier. Furthermore a net of dictionaries were employed to increase the reliability and precision of the system output. Finally, the Iran-Shahr database is utilized to evaluate the system performance. Comparing the results of the proposed method and other previous methods, proves that a less sensitivity has been achieved by the proposed method about handwriting variations.

PERSIAN specific Attribute for HANDWRITTEN Image WORD spotting and RECOGNITION by Label-EmbeddingOCR methods cannot handle degraded printed documents and HANDWRITTEN documents، WORD spotting is an alternative way to search for query WORDs. Nowadays، efficient Attribute-based Content Based Image Retrieval (CBIR) are developed for Image documents. In this paper new attributes، based on PERSIAN writing Style are introduced for PERSIAN WORD spotting. We compared and analyze our methods against the other state of the art methods. In addition، the proposed method can handle both Image and string queries، also it can be employed as a WORD RECOGNITION system. Experiments on two standard datasets، Farsa and Iranshahr، shows reasonable results.

In this paper we address the issue of recognizing Farsi HANDWRITTEN WORDs. Two types fo gradient features are extracted from a sliding vertical stripe which sweeps across a WORD image. These are directional and intensity gradient features. The feature vector extracted from each stripe is then coded using the Self Organizing Map (SOM). In this method each WORD is modeled using the discrete Hidden Markov Model (HMM). To evaluate the performance of the proposed method, FARSA dataset has been used. The experimental results show that the proposed system, applying directional gradient features, has achieved the RECOGNITION rate of 69.07% and outperformed all other existing methods.

In this paper, a new method is proposed for offline HANDWRITTEN PERSIAN WORDs RECOGNITION. The proposed method introduces the Centroid Sequence Freeman Chain Code (CSFCC) as a new and powerful feature along with the use of morphological features and an optimize support vector machine (SVM) classifier. A conbination of particle swarm optimization (PSO) and gravitational search algorithm (GSA), abbreviated to PSOGSA, has been employed to optimaze the SVM classifier. In the proposed method, all the connected components of a WORD are detected and combined with each other. For this purpose, a pictorial dictionary of asymptomatic subWORDs has been made. In addition, a database has been created to include the positions of asymptomatic subWORDs in order to narrow down the search space and increase the speed and improve the RECOGNITION accuracy. Based on the position of a subWORD in a WORD, it is more likely to make the right decision and detect the subWORD, accurately. The proposed method was implemented on the Iranshahr Database, containing nearly 17000 images of HANDWRITTEN names of 503 cities of Iran. The resultant RECOGNITION accuracy is 89% in the expriments, which shows the capability of the proposed method and improving the results, compared to the other well-known methods.

HANDWRITTEN WORD RECOGNITION (HWR) is very important in document analysis and retrieval. In this paper, an off-line HANDWRITTEN RECOGNITION system for PERSIAN manuscript is introduced.For feature extraction, SIFT descriptors extracted densely from the block of WORD image and enriched by appending the normalized x and y coordinates and the scale they were extracted at. Linear discriminate analysis (LDA) is used for feature reduction. All WORDs in the dictionary were hierarchically clustered by ISOCLUSE algorithm. In order to recognize the WORD images, multiple-class and two-class SVM classifiers methods were used. The experimental results showed a better performance in terms of speed and precision of two-class SVM method on the Iranshahr data set. The accuracy of proposed system by select 5 top cluster is shown 93.37% by 76.65% reduction of lexicon.

In this paper, a model of spoken WORD RECOGNITION is proposed. This model is particularly concerned with extraction of cues from the signal leading to a specification of a WORD in terms of bundles of distinctive features, which are assumed to be the building blocks of WORDs. In the model proposed, auditory input is chunked into a set of successive time slices. It is assumed that the derivation of the underlying WORD pattern proceeds in three layers: Features, phonemes, WORDs. The feature layer has a complete set of feature detectors at every time slice. In this layer, the detection of the underlying pattern of distinctive features from the speech signal proceeds 'in three steps. In the first step, numerical values for features are obtained measuring acoustic attributes in each time slice. The acoustic attributes are either acoustic landmarks corresponding to articulator-free features which are identified, based on amplitude changes in various energy bands, or acoustic cues in the vicinity of the landmarks corresponding to articulator-bound features. Continuous perceptual feature values are, then processed into a much more structured representation, namely phonological surface structure. This is carried out in Perception Grammar as suggested by Boersma (1998). In the third step, a further processing is carried out. to turn the discrete representation into an abstract one yielding the underlying pattern of distinctive features. The next layer of the model has a complete set of phoneme detectors for every three time slices, but each set spans six time slices so the sets overlap. This means that the detection of adjacent phonemes will also overlap; this is supposed to simulate coarticulation The top layer has a complete set of WORD detector centered on every three time slices; again, the sets overlap, the number of time slices per WORD detector is variable because it depends on the length of each individual WORD.