Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradientbased learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of two dimensional (2-D) shapes, are shown to outperform all other techniques.Real-life document recognition systems are composed of multiple modules including field extraction, segmentation, recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN's), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure.Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks.A graph transformer network for reading a bank check is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal checks. It is deployed commercially and reads several million checks per day.

Traditional classification approaches generalize poorly on image classification tasks, because of the high dimensionality of the feature space. This paper shows that support vector machines (SVM's) can generalize well on difficult image classification problems where the only features are high dimensional histograms. Heavy-tailed RBF kernels of the form K(x, y) = e(-rho)Sigma(i)/xia-yia/b with a < or = 1 and b < or = 2 are evaluated on the classification of images extracted from the Corel stock photo collection and shown to far outperform traditional polynomial or Gaussian radial basis function (RBF) kernels. Moreover, we observed that a simple remapping of the input x(i)-->x(i)(a) improves the performance of linear SVM's to such an extend that it makes them, for this problem, a valid alternative to RBF kernels.

No abstract

No abstract

We present a new image compression technique called \DjVu " that is speci cally geared towards the compression of high-resolution, high-quality images of scanned documents in color. This enables fast transmission of document images over low-speed connections, while faithfully reproducing the visual aspect of the document, including color, fonts, pictures, and paper texture. The DjVu compressor separates the text and drawings, which needs a high spatial resolution, from the pictures and backgrounds, which are smoother and can be coded at a lower spatial resolution. Then, several novel techniques are used to maximize the compression ratio: the bi-level foreground image is encoded with AT&T's proposal to the new JBIG2 fax standard, and a new wavelet-based compression method is used for the backgrounds and pictures. Both techniques use a new adaptive binary arithmetic coder called the Z-coder. A typical magazine page in color at 300dpi can be compressed down to betwe e n 4 0 t o 6 0 K B , a p p r o ximately 5 to 10 times better than JPEG for a similar level of subjective quality. A real-time, memory e cient v ersion of the decoder was implemented, and is available as a plug-in for popular web browsers.