Two Dimensional Principal Component Analysis for Online Tamil Character Recognition

This paper presents a new application of two dimensional Principal Component Analysis (2DPCA) to the problem of online character recognition in Tamil Script. A novel set of features employing polynomial fits and quartiles in combination with conventional features are derived for each sample point of the Tamil character obtained after smoothing and resampling. These are stacked to form a matrix, using which a covariance matrix is constructed. A subset of the eigenvectors of the covariance matrix is employed to get the features in the reduced sub space. Each character is modeled as a separate subspace and a modified form of the Mahalanobis distance is derived to classify a given test character. Results indicate that the recognition accuracy using the 2DPCA scheme shows an approximate 3% improvement over the conventional PCA technique.

Compiler-Directed Frequency and Voltage Scaling for a Multiple Clock Domain

Multiple Clock Domain processors provide an attractive solution to the increasingly challenging problems of clock distribution and power dissipation. They allow their chips to be partitioned into different clock domains, and each domain’s frequency (voltage) to be independently configured. This flexibility adds new dimensions to the Dynamic Voltage and Frequency Scaling problem, while providing better scope for saving energy and meeting performance demands. In this paper, we propose a compiler directed approach for MCD-DVFS. We build a formal petri net based program performance model, parameterized by settings of microarchitectural components and resource configurations, and integrate it with our compiler passes for frequency selection.Our model estimates the performance impact of a frequency setting, unlike the existing best techniques which rely on weaker indicators of domain performance such as queue occupancies(used by online methods) and slack manifestation for a particular frequency setting (software based methods).We evaluate our method with subsets of SPECFP2000,Mediabench and Mibench benchmarks. Our mean energy savings is 60.39% (versus 33.91% of the best software technique)in a memory constrained system for cache miss dominated benchmarks, and we meet the performance demands.Our ED2 improves by 22.11% (versus 18.34%) for other benchmarks. For a CPU with restricted frequency settings, our energy consumption is within 4.69% of the optimal.

Exploring use of climate information for optimal design of water resources infrastructure

We conducted surveys of fire and fuels managers at local, regional, and national levels to gain insights into decision processes and information flows in wildfire management. Survey results in the form of fire managers’ decision calendars show how climate information needs vary seasonally, over space, and through the organizational network, and help determine optimal points for introducing climate information and forecasts into decision processes. We identified opportunities to use climate information in fire management, including seasonal to interannual climate forecasts at all organizational levels, to improve the targeting of fuels treatments and prescribed burns, the positioning and movement of initial attack resources, and staffing and budgeting decisions. Longer-term (5–10 years) outlooks also could be useful at the national level in setting budget and research priorities. We discuss these opportunities and examine the kinds of organizational changes that could facilitate effective use of existing climate information and climate forecast capabilities.

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 Handwritten Kannada Word Recognizer with Unrestricted Vocabulary

In this paper, we propose a novel heuristic approach to segment recognizable symbols from online Kannada word data and perform recognition of the entire word. Two different estimates of first derivative are extracted from the preprocessed stroke groups and used as features for classification. Estimate 2 proved better resulting in 88% accuracy, which is 3% more than that achieved with estimate 1. Classification is performed by statistical dynamic space warping (SDSW) classifier which uses X, Y co-ordinates and their first derivatives as features. Classifier is trained with data from 40 writers. 295 classes are handled covering Kannada aksharas, with Kannada numerals, Indo-Arabic numerals, punctuations and other special symbols like $ and #. Classification accuracies obtained are 88% at the akshara level and 80% at the word level, which shows the scope for further improvement in segmentation algorithm

Mechanistic insights into type III restriction enzymes

Type III restriction-modification (R-M) enzymes need to interact with two separate unmethylated DNA sequences in indirectly repeated, head-to-head orientations for efficient cleavage to occur at a defined location next to only one of the two sites. However, cleavage of sites that are not in head-to-head orientation have been observed to occur under certain reaction conditions in vitro. ATP hydrolysis is required for the long-distance communication between the sites prior to cleavage. Type III R-M enzymes comprise two subunits, Res and Mod that form a homodimeric Mod(2) and a heterotetrameric Res(2)Mod(2) complex. The Mod subunit in M-2 or R2M2 complex recognizes and methylates DNA while the Res subunit in R2M2 complex is responsible for ATP hydrolysis, DNA translocation and cleavage. A vast majority of biochemical studies on Type III R-M enzymes have been undertaken using two closely related enzymes, EcoP1I and EcoP15I. Divergent opinions about how the long-distance interaction between the recognition sites exist and at least three mechanistic models based on 1D- diffusion and/or 3D-DNA looping have been proposed.

Differential Diffractive Reflectance Modulation Sensing

We fabricated a reflectance based sensor which relies on the diffraction pattern generated from a bio-microarray where an underlying thin film structure enhances the diffracted intensity from molecular layers. The zero order diffraction represents the background signal and the higher orders represent the phase difference between the array elements and the background. By taking the differential ratio of the first and zero order diffraction signals we get a quantitative measure of molecular binding while simultaneously rejecting common mode fluctuations. We improved the signal-to-noise ratio by an order of magnitude with this ratiometric approach compared to conventional single channel detection. In addition, we use a lithography based approach for fabricating microarrays which results in spot sizes as small as 5 micron diameter unlike the 100 micron spots from inkjet printing and is therefore capable of a high degree of multiplexing. We will describe the real-time measurement of adsorption of charged polymers and bulk refractometry using this technique. The lack of moving parts for point scanning of the microarray and the differential ratiometric measurements using diffracted orders from the same probe beam allows us to make real-time measurements in spite of noise arising from thermal or mechanical fluctuations in the fluid sample above the sensor surface. Further, the lack of moving parts leads to considerable simplification in the readout hardware permitting the use of this technique in compact point of care sensors.