OCR systems for printed documents typically require large numbers of font styles and character models to work well. When given an unseen font, performance degrades even in the absence of noise. In this paper, we perform OCR in an unsupervised fashion without using any character models by using a cryptogram decoding algorithm. We present results on real and artificial OCR data.

The fundamental task facing Optical Character Recognition (OCR) systems involves the conversion of input document images into corresponding sequences of symbolic character codes. Traditionally, this has been accomplished in a bottom-up fashion: the image of each symbol is isolated, then classified based on its pixel intensities. While such shape-based classifiers are initially trained on a wide array of fonts, they still tend to perform poorly when faced with novel glyph shapes. In this thesis, we attempt to bypass this problem by pursuing a top-down “codebreaking ” approach. We assume no a priori knowledge of character shape, instead relying on statistical information and language constraints to determine an appropriate character mapping. We introduce and contrast three new top-down approaches, and present experimental results on several real and synthetic datasets. Given sufficient amounts of data, our font and shape independent approaches are shown to perform about as well as shape-based classifiers. ii Acknowledgements First and foremost, I would like to thank my supervisor Sam Roweis for his tireless sup-

We consider models for which it is important, early in processing, to estimate some variables with high precision, but perhaps at relatively low rates of recall. If some variables can be identified with near certainty, then they can be conditioned upon, allowing further inference to be done efficiently. Specifically, we consider optical character recognition (OCR) systems that can be bootstrapped by identifying a subset of correctly translated document words with very high precision. This “clean set ” is subsequently used as document-specific training data. While many current OCR systems produce measures of confidence for the identity of each letter or word, thresholding these confidence values, even at very high values, still produces some errors. We introduce a novel technique for identifying a set of correct words with very high precision. Rather than estimating posterior probabilities, we bound the probability that any given word is incorrect under very general assumptions, using an approximate worst case analysis. As a result, the parameters of the model are nearly irrelevant, and we are able to identify a subset of words, even in noisy documents, of which we are highly confident. On our set of 10 documents, we are able to identify about 6 % of the words on average without making a single error. This ability to produce word lists with very high precision allows us to use a family of models which depends upon such clean word lists. 1

Abstract — The keystream reuse problem in case of textual data has been the focus of cryptanalysts for quite some time now. This paper presents the use of hidden markov models based speech recognition approach to cryptanalysis of stream ciphered digitized speech in a keystream reuse situation. In this paper, we show that how an adversary can automatically recover the digitized speech signals encrypted under the same keystream. The technique is flexible enough to incorporate all modern speech coding schemes and all languages for which the speech recognition techniques exist. The technique is simple and efficient and can be practically employed with the existing HMM based probabilistic speech recognition techniques with some modification in the training (pre-computation) and/or the maximum likelihood decoding procedure. The simulation experiments, though preliminary, showed promising initial results by recognizing about 80 percent correct phoneme pairs encrypted by the same keystream. Index Terms — cryptanalysis, hidden markov model, keystream reuse, speech recognition, stream cipher. I.

Abstract — Keystream reuse also known as the “two time pad” problem in case of stream ciphered data has been the focus of cryptanalysts for several decades. All heuristics presented so far assume the underlying plaintext to be uncompressed text based data encoded through conventional encoding mechanisms such as ASCII Coding. This paper presents the use of hidden Markov model (HMM) based automatic speech recognition (ASR) approach to cryptanalysis of stream-ciphered waveform-encoded speech in a keystream reuse situation. We present that an adversary can automatically recover the digitized speech signals from their plaintext XORs obtained from two different speech signals stream ciphered with the same keystream. The proposed technique can be practically employed with the existing HMM based probabilistic speech recognition techniques with some modification in the selection of HMMs, their training and the maximum likelihood decoding procedures. Simulation experiments using such modified speech recognition tools have been presented. Index Terms — cryptanalysis, keystream reuse, speech coding, stream cipher, two time pad. I.

Despite ubiquitous claims that optical character recognition (OCR) is a “solved problem, ” many categories of documents continue to break modern OCR software such as documents with moderate degradation or unusual fonts. Many approaches rely on pre-computed or stored character models, but these are vulnerable to cases when the font of a particular document was not part of the training set, or when there is so much noise in a document that the font model becomes weak. To address these difficult cases, we present a form of iterative contextual modeling that learns character models directly from the document it is trying to recognize. We use these learned models both to segment the characters and to recognize them in an incremental, iterative process. We present results comparable to those of a commercial OCR system on a subset of characters from a difficult test document. 1.

Despite ubiquitous claims that optical character recognition (OCR) is a “solved problem, ” many categories of documents continue to break modern OCR software such as documents with moderate degradation or unusual fonts. Many approaches rely on pre-computed or stored character models, but these are vulnerable to cases when the font of a particular document was not part of the training set, or when there is so much noise in a document that the font model becomes weak. To address these difficult cases, we present a form of iterative contextual modeling that learns character models directly from the document it is trying to recognize. We use these learned models both to segment the characters and to recognize them in an incremental, iterative process. We present results comparable to those of a commercial OCR system on a subset of characters from a difficult test document. 1.

Abstract Despite ubiquitous claims that optical character recognition (OCR) is a “solved problem, ” many categories of documents continue to break modern OCR software such as documents with moderate degradation or unusual fonts. Many approaches rely on pre-computed or stored character models, but these are vulnerable to cases when the font of a particular document was not part of the training set or when there is so much noise in a document that the font model becomes weak. To address these difficult cases, we present a form of iterative contextual modeling that learns character models directly from the document it is trying to recognize. We use these learned models both to segment the characters and to recognize them in an incremental, iterative process. We present results comparable with those of a commercial OCR system on a subset of characters from a difficult test document in both English and Greek.