Recognising lines of unconstrained handwritten text is a challenging task. The difficulty of segmenting cursive or overlapping characters, combined with the need to exploit surrounding context, has led to low recognition rates for even the best current recognisers. Most recent progress in the field has been made either through improved preprocessing, or through advances in language modelling. Relatively little work has been done on the basic recognition algorithms. Indeed, most systems rely on the same hidden Markov models that have been used for decades in speech and handwriting recognition, despite their well-known shortcomings. This paper proposes an alternative approach based on a novel type of recurrent neural network, specifically designed for sequence labelling tasks where the data is hard to segment and contains long range, bidirectional interdependencies. In experiments on two large unconstrained handwriting databases, our approach achieves word recognition accuracies of 79.7 % on online data and 74.1 % on offline data, significantly outperforming a state-of-the-art HMM-based system. In addition, we demonstrate the network’s robustness to lexicon size, measure the individual influence of its hidden layers, and analyse its use of context. Lastly we provide an in depth discussion of the differences between the network and HMMs, suggesting reasons for the network’s superior performance.

Abstract. In this paper we present a method for synthesizing English handwritten textlines from ASCII transcriptions. The method is based on templates of characters and the Delta LogNormal model of handwriting generation. To generate a textline, first a static image of the textline is built by concatenating perturbed versions of the character templates. Then strokes and corresponding virtual targets are extracted and randomly perturbed, and finally the textline is drawn using overlapping strokes and delta-lognormal velocity profiles in accordance with the Delta LogNormal theory. The generated textlines are used as training data for a hidden Markov model based off-line handwritten textline recognizer. First results show that adding such generated textlines to the natural training set may be beneficial.

In this paper we introduce a new connectionist approach to on-line handwriting recognition and address in particular the problem of recognizing handwritten whiteboard notes. The approach uses a bidirectional recurrent neural network with the long short-term memory architecture. We use a recently introduced objective function, known as Connectionist Temporal Classification (CTC), that directly trains the network to label unsegmented sequence data. Our new system achieves a word recognition rate of 74.0 %, compared with 65.4 % using a previously developed HMMbased recognition system. 1.

In this paper we propose a novel approach to the detection of on-line handwritten text lines based on dynamic programming. We try to find the paths with the minimum cost between two consecutive text lines. Most steps of the proposed algorithm are based on off-line information. Hence the method can also be applied to off-line documents after a few minor changes. In our experiments we show that this dynamic programming based approach is better than a common on-line segmentation procedure. 1.

This paper presents writer-independent offline handwritten character recognition for Ethiopic script. The recognition is based on the characteristics of primitive strokes that make up characters. The spatial relationships of primitives whose combinations form complex structures of Ethiopic characters are used as a basis for recognition. Although this approach efficiently recognizes properly written characters, the recognition rate drops for characters where the spatial relationships of their primitives could not be drawn. This happens mostly when the connections between primitives are not properly written, which is a common case in handwriting. To complement the recognition, we classify characters based on the characteristics of their primitives, resulting in grouping of characters in a five-dimensional space. Once the type of characters is identified, recognition can be achieved with a minimal set of information from their spatial relationships. A comprehensive database is also developed to standardize the evaluation of research works on offline Ethiopic handwriting recognition systems. Our proposed system is tested is with the database and experimental results are reported. Keywords: Ethiopic, Handwriting Recognition, Database. 1.

This paper describes an offline handwriting recognition system for Amharic words based on lexicon. The system computes direction fields of scanned handwritten documents, from which pseudocharacters are segmented. The pseudo-characters are organized based on their proximity and direction to form text lines. Words are then segmented by analyzing the relative gap between subsequent pseudo-characters in text lines. For each segmented word image, the structural characteristics of pseudocharacters are syntactically analyzed to predict a set of plausible characters forming the word. The most likelihood word is finally selected among candidates by matching against the lexicon. The system is tested by a database of unconstrained handwritten Amharic documents collected from various sources. The lexicon is prepared from words appearing in the collected database. Experimental results are reported. 1.

cursive word representation based on

Free-style sketching is more natural than drawing a sketch using mouse and palette based tool. A number of sketching systems have been developed but not much attention is paid towards friendliness of the system. We present here an efficient Sketching system MAHII for sketching the strokes from multiple domains. The proposed paper demonstrates how a sketching system can be made user friendly and user interactive too. We also present the architecture of our system that solves many flaws encountered in the previously built systems. The interface should be such that it minimizes the gap between a human being and a machine. Our system provides a robust architecture that proves to be efficient. It also allows the user to sketch the diagram freely. And at the same time it edits the sketches as well as recognizes the sketches efficiently. This paper represents more robust and natural environment to the user. The MAHII provides the user an interface that is built in accordance with a novice user. MAHII provides free style sketching, recognizes the sketch produced by the user with accurate results. Keywords: MAHII, sketch, interface, stroke, multi-domain, Heuristics. 1.

In this paper, an improvement of a 2D stochastic model based handwritten entity recognition system is described. To model the handwriting considered as being a two dimensional signal, a context based, segmentation-free Hidden Markov Model (HMM) recognition system was used. The baseline approach combines a Markov Random Field (MRF) and a HMM so-called Non-Symmetric Half Plane Hidden Markov Model (NSHP-HMM). To improve the results performed by this baseline system operating just on low-level pixel information an extension of the NSHP-HMM is proposed. The mechanism allows to extend the observations of the NSHP-HMM by implanting structural information in the system. At present, the accuracy of the system on the SRTP 1 French postal check database is 87.52 % while for the handwritten Bangla city names is 86.80%. The gain using this structural information for the SRTP dataset is 1.57%. 1.

This chapter reports our knowledge about the analysis and recognition of scanned administrative documents. Regarding essentially the administrative paper flow with new and continuous arrivals, all the conventional techniques reserved to static databases modeling and recognition are doomed to failure. For this purpose, a new technique based on the experience was investigated giving very promising results. This technique is related to the case-based reasoning already used in data mining and various problems of machine learning. After the presentation of the context related to the administrative document flow and its requirements in a real time processing, we present a case based reasonning for invoice processing. The case corresponds to the co-existence of a problem and its solution. The problem in an invoice corresponds to a local structure such as the keywords of an address or the line patterns in the amounts table, while the solution is related to their content. This problem is then compared to a document case base using graph probing. For this purpose, we proposed an improvement of an already existing neural network called Incremental Growing Neural Gas 1