Creation of a Huge Annotated Database for Tamil and Kannada OHR

This paper describes the efforts at MILE lab, IISc, to create a 100,000-word database each in Kannada and Tamil for the design and development of Online Handwritten Recognition. It has been collected from over 600 users in order to capture the variations in writing style. We describe features of the scripts and how the number of symbols were reduced to be able to effectively train the data for recognition. The list of words include all the characters, Kannada and Indo-Arabic numerals, punctuations and other symbols. A semi-automated tool for the annotation of data from stroke to word level is used. It segments each word into stroke groups and also acts as a validation mechanism for segmentation. The tool displays the stroke, stroke groups and aksharas of a word and hence can be used to study the various styles of writing, delayed strokes and for assigning quality tags to the words. The tool is currently being used for annotating Tamil and Kannada data. The output is stored in a standard XML format.

Online Handwriting Recognition of Tamil Script Using Fractal Geometry

We present a fractal coding method to recognize online handwritten Tamil characters and propose a novel technique to increase the efficiency in terms of time while coding and decoding. This technique exploits the redundancy in data, thereby achieving better compression and usage of lesser memory. It also reduces the encoding time and causes little distortion during reconstruction. Experiments have been conducted to use these fractal codes to classify the online handwritten Tamil characters from the IWFHR 2006 competition dataset. In one approach, we use fractal coding and decoding process. A recognition accuracy of 90% has been achieved by using DTW for distortion evaluation during classification and encoding processes as compared to 78% using nearest neighbor classifier. In other experiments, we use the fractal code, fractal dimensions and features derived from fractal codes as features in separate classifiers. While the fractal code is successful as a feature, the other two features are not able to capture the wide within-class variations.

Comparison of HMM and SDTW for Tamil handwritten character recognition

In this paper, we compare the experimental results for Tamil online handwritten character recognition using HMM and Statistical Dynamic Time Warping (SDTW) as classifiers. HMM was used for a 156-class problem. Different feature sets and values for the HMM states & mixtures were tried and the best combination was found to be 16 states & 14 mixtures, giving an accuracy of 85%. The features used in this combination were retained and a SDTW model with 20 states and single Gaussian was used as classifier. Also, the symbol set was increased to include numerals, punctuation marks and special symbols like $, & and #, taking the number of classes to 188. It was found that, with a small addition to the feature set, this simple SDTW classifier performed on par with the more complicated HMM model, giving an accuracy of 84%. Mixture density estimation computations was reduced by 11 times. The recognition is writer independent, as the dataset used is quite large, with a variety of handwriting styles.

Computational Biology and Language

Current scientific research is characterized by increasing specialization, accumulating knowledge at a high speed due to parallel advances in a multitude of sub-disciplines. Recent estimates suggest that human knowledge doubles every two to three years – and with the advances in information and communication technologies, this wide body of scientific knowledge is available to anyone, anywhere, anytime. This may also be referred to as ambient intelligence – an environment characterized by plentiful and available knowledge. The bottleneck in utilizing this knowledge for specific applications is not accessing but assimilating the information and transforming it to suit the needs for a specific application. The increasingly specialized areas of scientific research often have the common goal of converting data into insight allowing the identification of solutions to scientific problems. Due to this common goal, there are strong parallels between different areas of applications that can be exploited and used to cross-fertilize different disciplines. For example, the same fundamental statistical methods are used extensively in speech and language processing, in materials science applications, in visual processing and in biomedicine. Each sub-discipline has found its own specialized methodologies making these statistical methods successful to the given application. The unification of specialized areas is possible because many different problems can share strong analogies, making the theories developed for one problem applicable to other areas of research. It is the goal of this paper to demonstrate the utility of merging two disparate areas of applications to advance scientific research. The merging process requires cross-disciplinary collaboration to allow maximal exploitation of advances in one sub-discipline for that of another. We will demonstrate this general concept with the specific example of merging language technologies and computational biology.

Time Series Classification for Online Tamil Handwritten Character Recognition – A Kernel Based Approach

In this paper, we consider the problem of time series classification. Using piecewise linear interpolation various novel kernels are obtained which can be used with Support vector machines for designing classifiers capable of deciding the class of a given time series. The approach is general and is applicable in many scenarios. We apply the method to the task of Online Tamil handwritten character recognition with promising results.

Language identification using parallel sub-word recognition

Parallel sub-word recognition (PSWR) is a new model that has been proposed for language identification (LID) which does not need elaborate phonetic labeling of the speech data in a foreign language. The new approach performs a front-end tokenization in terms of sub-word units which are designed by automatic segmentation, segment clustering and segment HMM modeling. We develop PSWR based LID in a framework similar to the parallel phone recognition (PPR) approach in the literature. This includes a front-end tokenizer and a back-end language model, for each language to be identified. Considering various combinations of the statistical evaluation scores, it is found that PSWR can perform as well as PPR, even with broad acoustic sub-word tokenization, thus making it an efficient alternative to the PPR system.

Suite of tools for statistical N-gram language modeling for pattern mining in whole genome sequences

Genome sequences contain a number of patterns that have biomedical significance. Repetitive sequences of various kinds are a primary component of most of the genomic sequence patterns. We extended the suffix-array based Biological Language Modeling Toolkit to compute n-gram frequencies as well as n-gram language-model based perplexity in windows over the whole genome sequence to find biologically relevant patterns. We present the suite of tools and their application for analysis on whole human genome sequence.

Estimation of design basis earthquake using region-specific M-max, for the NPP site at Kalpakkam, Tamil Nadu, India

The objective of the paper is to estimate Safe Shutdown Earthquake (SSE) and Operating/Design Basis Earthquake (OBE/DBE) for the Nuclear Power Plant (NPP) site located at Kalpakkam, Tamil Nadu, India. The NPP is located at 12.558 degrees N, 80.175 degrees E and a 500 km circular area around NPP site is considered as `seismic study area' based on past regional earthquake damage distribution. The geology, seismicity and seismotectonics of the study area are studied and the seismotectonic map is prepared showing the seismic sources and the past earthquakes. Earthquake data gathered from many literatures are homogenized and declustered to form a complete earthquake catalogue for the seismic study area. The conventional maximum magnitude of each source is estimated considering the maximum observed magnitude (M-max(obs)) and/or the addition of 0.3 to 0.5 to M-max(obs). In this study maximum earthquake magnitude has been estimated by establishing a region's rupture character based on source length and associated M-max(obs). A final source-specific M-max is selected from the three M-max values by following the logical criteria. To estimate hazard at the NPP site, ten Ground-Motion Prediction Equations (GMPEs) valid for the study area are considered. These GMPEs are ranked based on Log-Likelihood (LLH) values. Top five GMPEs are considered to estimate the peak ground acceleration (PGA) for the site. Maximum PGA is obtained from three faults and named as vulnerable sources to decide the magnitudes of OBE and SSE. The average and normalized site specific response spectrum is prepared considering three vulnerable sources and further used to establish site-specific design spectrum at NPP site.

PolyGLoT: A Polyhedral Loop Transformation Framework for a Graphical Dataflow Language

Polyhedral techniques for program transformation are now used in several proprietary and open source compilers. However, most of the research on polyhedral compilation has focused on imperative languages such as C, where the computation is specified in terms of statements with zero or more nested loops and other control structures around them. Graphical dataflow languages, where there is no notion of statements or a schedule specifying their relative execution order, have so far not been studied using a powerful transformation or optimization approach. The execution semantics and referential transparency of dataflow languages impose a different set of challenges. In this paper, we attempt to bridge this gap by presenting techniques that can be used to extract polyhedral representation from dataflow programs and to synthesize them from their equivalent polyhedral representation. We then describe PolyGLoT, a framework for automatic transformation of dataflow programs which we built using our techniques and other popular research tools such as Clan and Pluto. For the purpose of experimental evaluation, we used our tools to compile LabVIEW, one of the most widely used dataflow programming languages. Results show that dataflow programs transformed using our framework are able to outperform those compiled otherwise by up to a factor of seventeen, with a mean speed-up of 2.30x while running on an 8-core Intel system.

Coupling geomorphology parameters with lithology for micro-level groundwater resource assessment: a case study from semi-arid hard rock terrain in Tamil Nadu, India

The standard procedure of groundwater resource estimation in India till date is based on the specific yield parameters of each rock type (lithology) derived through pumping test analysis. Using the change in groundwater level, specific yield, and area of influence, groundwater storage change could be estimated. However, terrain conditions in the form of geomorphological variations have an important bearing on the net groundwater recharge. In this study, an attempt was made to use both lithology and geomorphology as input variables to estimate the recharge from different sources in each lithology unit influenced by the geomorphic conditions (lith-geom), season wise separately. The study provided a methodological approach for an evaluation of groundwater in a semi-arid hard rock terrain in Tirunelveli, Tamil Nadu, India. While characterizing the gneissic rock, it was found that the geomorphologic variations in the gneissic rock due to weathering and deposition behaved differently with respect to aquifer recharge. The three different geomorphic units identified in gneissic rock (pediplain shallow weathered (PPS), pediplain moderate weathered (PPM), and buried pediplain moderate (BPM)) showed a significant variation in recharge conditions among themselves. It was found from the study that Peninsular gneiss gives a net recharge value of 0.13 m/year/unit area when considered as a single unit w.r.t. lithology, whereas the same area considered with lith-geom classes gives recharge values between 0.1 and 0.41 m/year presenting a different assessment. It is also found from this study that the stage of development (SOD) for each lith-geom unit in Peninsular gneiss varies from 168 to 230 %, whereas the SOD is 223 % for the lithology as a single unit.

Translation Induction on Indian Language Corpora Using Translingual Themes from Other Languages

Identifying translations from comparable corpora is a well-known problem with several applications, e.g. dictionary creation in resource-scarce languages. Scarcity of high quality corpora, especially in Indian languages, makes this problem hard, e.g. state-of-the-art techniques achieve a mean reciprocal rank (MRR) of 0.66 for English-Italian, and a mere 0.187 for Telugu-Kannada. There exist comparable corpora in many Indian languages with other ``auxiliary'' languages. We observe that translations have many topically related words in common in the auxiliary language. To model this, we define the notion of a translingual theme, a set of topically related words from auxiliary language corpora, and present a probabilistic framework for translation induction. Extensive experiments on 35 comparable corpora using English and French as auxiliary languages show that this approach can yield dramatic improvements in performance (e.g. MRR improves by 124% to 0.419 for Telugu-Kannada). A user study on WikiTSu, a system for cross-lingual Wikipedia title suggestion that uses our approach, shows a 20% improvement in the quality of titles suggested.

Falcon: A Graph Manipulation Language for Heterogeneous Systems

Graph algorithms have been shown to possess enough parallelism to keep several computing resources busy-even hundreds of cores on a GPU. Unfortunately, tuning their implementation for efficient execution on a particular hardware configuration of heterogeneous systems consisting of multicore CPUs and GPUs is challenging, time consuming, and error prone. To address these issues, we propose a domain-specific language (DSL), Falcon, for implementing graph algorithms that (i) abstracts the hardware, (ii) provides constructs to write explicitly parallel programs at a higher level, and (iii) can work with general algorithms that may change the graph structure (morph algorithms). We illustrate the usage of our DSL to implement local computation algorithms (that do not change the graph structure) and morph algorithms such as Delaunay mesh refinement, survey propagation, and dynamic SSSP on GPU and multicore CPUs. Using a set of benchmark graphs, we illustrate that the generated code performs close to the state-of-the-art hand-tuned implementations.

Discriminative language model adaptation for Mandarin broadcast speech transcription and translation

This paper investigates unsupervised test-time adaptation of language models (LM) using discriminative methods for a Mandarin broadcast speech transcription and translation task. A standard approach to adapt interpolated language models to is to optimize the component weights by minimizing the perplexity on supervision data. This is a widely made approximation for language modeling in automatic speech recognition (ASR) systems. For speech translation tasks, it is unclear whether a strong correlation still exists between perplexity and various forms of error cost functions in recognition and translation stages. The proposed minimum Bayes risk (MBR) based approach provides a flexible framework for unsupervised LM adaptation. It generalizes to a variety of forms of recognition and translation error metrics. LM adaptation is performed at the audio document level using either the character error rate (CER), or translation edit rate (TER) as the cost function. An efficient parameter estimation scheme using the extended Baum-Welch (EBW) algorithm is proposed. Experimental results on a state-of-the-art speech recognition and translation system are presented. The MBR adapted language models gave the best recognition and translation performance and reduced the TER score by up to 0.54% absolute. © 2007 IEEE.

Language model combination and adaptation using weighted finite state transducers

In speech recognition systems language model (LMs) are often constructed by training and combining multiple n-gram models. They can be either used to represent different genres or tasks found in diverse text sources, or capture stochastic properties of different linguistic symbol sequences, for example, syllables and words. Unsupervised LM adaptation may also be used to further improve robustness to varying styles or tasks. When using these techniques, extensive software changes are often required. In this paper an alternative and more general approach based on weighted finite state transducers (WFSTs) is investigated for LM combination and adaptation. As it is entirely based on well-defined WFST operations, minimum change to decoding tools is needed. A wide range of LM combination configurations can be flexibly supported. An efficient on-the-fly WFST decoding algorithm is also proposed. Significant error rate gains of 7.3% relative were obtained on a state-of-the-art broadcast audio recognition task using a history dependently adapted multi-level LM modelling both syllable and word sequences. ©2010 IEEE.