Recurrent neural networks are powerful sequence learners. They are able to incorporate context information in a flexible way, and are robust to localised distortions of the input data. These properties make them well suited to sequence labelling, where input sequences are transcribed with streams of labels. The aim of this thesis is to advance the state-of-the-art in supervised sequence labelling with recurrent networks. Its two main contributions are (1) a new type of output layer that allows recurrent networks to be trained directly for sequence labelling tasks where the alignment between the inputs and the labels is unknown, and (2) an extension of the long short-term memory network architecture to multidimensional data, such as images and video sequences.

https://mediatum.ub.tum.de/doc/673554/file.pdf

Supervised Sequence Labelling with Recurrent Neural Networks

Using recurrent neural network models for early detection of heart failure onset.

Recurrent neural networks (RNNs) with Long Short-Term memory cells currently hold the best known results in unconstrained handwriting recognition. We show that their performance can be greatly improved using dropout - a recently proposed regularization method for deep architectures. While previous works showed that dropout gave superior performance in the context of convolutional networks, it had never been applied to RNNs. In our approach, dropout is carefully used in the network so that it does not affect the recurrent connections, hence the power of RNNs in modeling sequences is preserved. Extensive experiments on a broad range of handwritten databases confirm the effectiveness of dropout on deep architectures even when the network mainly consists of recurrent and shared connections.

/pdf/dropout-improves-recurrent-neural-networks-for-handwriting-2di6lmbs26.pdf

Dropout Improves Recurrent Neural Networks for Handwriting Recognition

We present a Max-Pooling Convolutional Neural Network approach for supervised steel defect classification. On a classification task with 7 defects, collected from a real production line, an error rate of 7% is obtained. Compared to SVM classifiers trained on commonly used feature descriptors our best net performs at least two times better. Not only we do obtain much better results, but the proposed method also works directly on raw pixel intensities of detected and segmented steel defects, avoiding further time consuming and hard to optimize ad-hoc preprocessing.

/pdf/steel-defect-classification-with-max-pooling-convolutional-3zr7o1svd7.pdf

Steel defect classification with Max-Pooling Convolutional Neural Networks

Current state-of-the-art approaches to offline Handwritten Text Recognition extensively rely on Multidimensional Long Short-Term Memory networks However, these architectures come with quite an expensive computational cost, and we observe that they extract features visually similar to those of convolutional layers, which are computationally cheaper This suggests that the two-dimensional long-term dependencies, which are potentially modeled by multidimensional recurrent layers, may not be essential to achieve a good recognition accuracy, at least in the lower layers of the architecture In this work, an alternative model is explored that relies only on convolutional and one-dimensional recurrent layers that achieves better or equivalent results than those of the current state-of-the-art architecture, and runs significantly faster In addition, we observe that using random distortions during training as synthetic data augmentation dramatically improves the accuracy of our model Thus, are multidimensional recurrent layers really necessary for Handwritten Text Recognition? Probably not

Are Multidimensional Recurrent Layers Really Necessary for Handwritten Text Recognition

The paper presents the French RIMES (Reconnaissance et Indexation de donnees Manuscrites et de fac similES / Recognition and Indexing of handwritten documents and faxes) project, which is an evaluation campaign dedicated to automatic systems of recognition and indexing of handwritten letters such as those sent by postal mail or fax by individuals to companies or administrations. This project is intended to all the key players of the handwritten recognition and document analysis fields by proposing them to participate to evaluation tests, a first test having taken already place. In this framework, a new database of more than 5600 real mails completely annotated have been created as well as secondary databases of isolated characters, handwritten words (300 000 snippets) and logos. Beyond the evaluation, the other contribution of the RIMES project is to make available all these databases to the community after the last test

RIMES evaluation campaign for handwritten mail processing

This paper describes the handwriting recognition competitionheld at ICDAR 2009. This competition is based onthe RIMES-database, with French written text documents.These document are classified in three different categories,complete text pages, words, and isolated characters. Thisyear 10 systems were submitted for the handwritten recognitioncompetition on snippets of French words. The systemswere evaluated in three subtask depending of the sizes ofthe used dictionary. A comparison between different classificationand recognition systems show interesting results. Ashort description of the participating groups, their systems,and the results achieved are presented.

http://www.cvc.uab.es/icdar2009/papers/3725b398.pdf

ICDAR 2009 Handwriting Recognition Competition

This paper describes the French handwriting recognition competition held at ICDAR 2011. This competition is based on the RIMES-database composed of French written documents corresponding to letters sent by individuals to companies or administrations. Two tasks have been proposed this year : the first one consists in recognizing isolated snippets of words with the help of a given dictionary, the second one consists in recognizing blocks of words segmented into lines. This year 9 systems were submitted for the different competition subtasks. A comparison between different classification and recognition systems show interesting results. A short description of the participating groups, their systems, and the results achieved are presented.

http://www.quaero.org/media/files/bibliographie/grosicki_french_handwriting_icdar2011.pdf

ICDAR 2011 - French Handwriting Recognition Competition

This paper presents the results of the second test phase of the RIMES evaluation campaign. The latter is the first large-scale evaluation campaign intended to all the key players of the handwritten recognition and document analysis communities. It proposes various tasks around recognition and indexing of handwritten letters such as those sent by postal mail or fax by individuals to companies or administrations. In this second evaluation test, automatic systems have been evaluated on three themes: layout analysis, handwriting recognition and writer identification. The databases used are part of the RIMES database of 5605 real mails completely annotated as well as secondary databases of isolated characters and handwritten words (250,000 snippets). The paper reports on protocols and gives the results obtained in the campaign.(RIMES : Reconnaissance et Indexation de donnees Manuscrites et de fac similES / Recognition and Indexing of handwritten documents and faxes)

http://www.cvc.uab.es/icdar2009/papers/3725a941.pdf

Results of the RIMES Evaluation Campaign for Handwritten Mail Processing

Many preprocessing techniques have been proposed for isolated word recognition. However, recently, recognition systems have dealt with text blocks and their compound text lines. In this paper, we propose a new preprocessing approach to efficiently correct baseline skew and fluctuations. Our approach is based on a sliding window within which the vertical position of the baseline is estimated. Segmentation of text lines into subparts is, thus, avoided. Experiments conducted on a large publicly available database (Rimes), with a BLSTM (bidirectional long short-term memory) recurrent neural network recognition system, show that our baseline correction approach highly improves performance.

Emmanuèle Grosicki

Papers

RIMES evaluation campaign for handwritten mail processing

ICDAR 2009 Handwriting Recognition Competition

ICDAR 2011 - French Handwriting Recognition Competition

Results of the RIMES Evaluation Campaign for Handwritten Mail Processing

New baseline correction algorithm for text-line recognition with bidirectional recurrent neural networks