Efficient classification using phrases generated by topic models

There are many popular models available for classification of documents like Naïve Bayes Classifier, k-Nearest Neighbors and Support Vector Machine. In all these cases, the representation is based on the “Bag of words” model. This model doesn't capture the actual semantic meaning of a word in a particular document. Semantics are better captured by proximity of words and their occurrence in the document. We propose a new “Bag of Phrases” model to capture this discriminative power of phrases for text classification. We present a novel algorithm to extract phrases from the corpus using the well known topic model, Latent Dirichlet Allocation(LDA), and to integrate them in vector space model for classification. Experiments show a better performance of classifiers with the new Bag of Phrases model against related representation models.

Relook into the historical progress and philosophy of Indian space exploration

A study of the history and philosophy of the contribution of India towards the exploration of space since antiquity provides interesting insights. The contributions are described during the three periods namely: (1) the ten millenniums from 10,000 BC with a twilight period up to 900 AD; (2) the ten centuries from 900 AD to 1900 AD; and (3) the ten decades from 1900 AD to 2000 AD; called mythological, medieval, and modern respectively. Some important events during the above periods provide a reference view of the progress. The Vedas during the mythological period and the Siddhantas during the medieval periods, which are based on astronomical observations, indicate that the Indian contribution preceded other cultures. But most Western historians ignore this fact time and again in spite of many proofs provided to the contrary. This chapter also shows that Indians had the proper scientific attitude of developing any physical theory through the triplet of mind, model, and measurements. It is this same triplet that forms the basis of the present day well known Kalman filter technique. Up to about 1500 BC the Indian contribution was leading but during foreign invasion and occupation it lagged and has been improving only after independence.

Subject independent human action recognition using spatio-depth information and meta-cognitive RBF network

In this paper, we present a machine learning approach for subject independent human action recognition using depth camera, emphasizing the importance of depth in recognition of actions. The proposed approach uses the flow information of all 3 dimensions to classify an action. In our approach, we have obtained the 2-D optical flow and used it along with the depth image to obtain the depth flow (Z motion vectors). The obtained flow captures the dynamics of the actions in space time. Feature vectors are obtained by averaging the 3-D motion over a grid laid over the silhouette in a hierarchical fashion. These hierarchical fine to coarse windows capture the motion dynamics of the object at various scales. The extracted features are used to train a Meta-cognitive Radial Basis Function Network (McRBFN) that uses a Projection Based Learning (PBL) algorithm, referred to as PBL-McRBFN, henceforth. PBL-McRBFN begins with zero hidden neurons and builds the network based on the best human learning strategy, namely, self-regulated learning in a meta-cognitive environment. When a sample is used for learning, PBLMcRBFN uses the sample overlapping conditions, and a projection based learning algorithm to estimate the parameters of the network. The performance of PBL-McRBFN is compared to that of a Support Vector Machine (SVM) and Extreme Learning Machine (ELM) classifiers with representation of every person and action in the training and testing datasets. Performance study shows that PBL-McRBFN outperforms these classifiers in recognizing actions in 3-D. Further, a subject-independent study is conducted by leave-one-subject-out strategy and its generalization performance is tested. It is observed from the subject-independent study that McRBFN is capable of generalizing actions accurately. The performance of the proposed approach is benchmarked with Video Analytics Lab (VAL) dataset and Berkeley Multimodal Human Action Database (MHAD). (C) 2013 Elsevier Ltd. All rights reserved.

Form and Function of Proteins. Historical Background and the Indian Effort in Macromolecular Crystallography

Global efforts in macromolecular crystallography started in the thirties of the last century. However, definitive results began to emerge only in the late fifties and the early sixties. India has a long tradition in crystallography. The country had a head start in theoretical and computational structural biology, thanks to the efforts of G.N. Ramachandran and his colleagues in the fifties and the sixties. However, macromolecular crystallography got off the ground in India only in the eighties, particularly after the Bangalore group received adequate support from the Department of Science and Technology under their Thrust Area Programme. The Bangalore centre was also identified as a national nucleus for the development of the area in the country. Since then work in the area has spread widely and is being carried out by several groups, mainly led by scientists trained at Bangalore or their descendents, in about thirty institutions in India. In addition to the Department of Science and Technology, the effort is now supported by other agencies like the Department of Biotechnology and the Council of Scientific and Industrial Research. The problems addressed by macromolecular crystallographers in India encompass almost all aspects of modern biology. Indian efforts in macromolecular crystallography have also become an important component of the international efforts in the area.

STUDENT'S-t MIXTURE MODEL BASED MULTI-INSTRUMENT RECOGNITION IN POLYPHONIC MUSIC

We address the problem of multi-instrument recognition in polyphonic music signals. Individual instruments are modeled within a stochastic framework using Student's-t Mixture Models (tMMs). We impose a mixture of these instrument models on the polyphonic signal model. No a priori knowledge is assumed about the number of instruments in the polyphony. The mixture weights are estimated in a latent variable framework from the polyphonic data using an Expectation Maximization (EM) algorithm, derived for the proposed approach. The weights are shown to indicate instrument activity. The output of the algorithm is an Instrument Activity Graph (IAG), using which, it is possible to find out the instruments that are active at a given time. An average F-ratio of 0 : 7 5 is obtained for polyphonies containing 2-5 instruments, on a experimental test set of 8 instruments: clarinet, flute, guitar, harp, mandolin, piano, trombone and violin.

Common recognition principles across diverse sequence and structural families of sialic acid binding proteins

Sialic acids form a large family of 9-carbon monosaccharides and are integral components of glycoconjugates. They are known to bind to a wide range of receptors belonging to diverse sequence families and fold classes and are key mediators in a plethora of cellular processes. Thus, it is of great interest to understand the features that give rise to such a recognition capability. Structural analyses using a non-redundant data set of known sialic acid binding proteins was carried out, which included exhaustive binding site comparisons and site alignments using in-house algorithms, followed by clustering and tree computation, which has led to derivation of sialic acid recognition principles. Although the proteins in the data set belong to several sequence and structure families, their binding sites could be grouped into only six types. Structural comparison of the binding sites indicates that all sites contain one or more different combinations of key structural features over a common scaffold. The six binding site types thus serve as structural motifs for recognizing sialic acid. Scanning the motifs against a non-redundant set of binding sites from PDB indicated the motifs to be specific for sialic acid recognition. Knowledge of determinants obtained from this study will be useful for detecting function in unknown proteins. As an example analysis, a genome-wide scan for the motifs in structures of Mycobacterium tuberculosis proteome identified 17 hits that contain combinations of the features, suggesting a possible function of sialic acid binding by these proteins.

Triarylborane-Appended New Triad and Tetrad: Chromogenic and Fluorogenic Anion Recognition

Facile synthesis of triad 3 and tetrad 4 incorporating -B(Mes)(2) (Mes = mesityl (2,4,6-trimethylphenyl)), boron dipyrromethene (BODIPY), and triphenylamine is reported. Introduction of two dissimilar acceptors (triarylborane and BODIPY) on a single donor resulted in two distinct intramolecular charge transfer processes (amine-to-borane and amine-to-BODIPY). The absorption and emission properties of the new triad and tetrad are highly dependent on individual building units. The nature of electronic communication among the individual fluorophore units has been comprehensively investigated and compared with building units. Compounds 3 and 4 showed chromogenic and fluorogenic responses for small anions such as fluoride and cyanide.

Dual Binding Site Assisted Chromogenic and Fluorogenic Recognition and Discrimination of Fluoride and Cyanide by a Peripherally Borylated Metalloporphyrin: Overcoming Anion Interference in Organoboron Based Sensors

Peripherally triarylborane decorated porphyrin (2) and its Zn(II) complex (3) have been synthesized. Compound 3 contains of two different Lewis acidic binding sites (Zn(II) and boron center). Unlike all previously known triarylborane based sensors, the optical responses of 3 toward fluoride and cyanide are distinctively different, thus enabling the discrimination of these two interfering anions. Metalloporphyrin 3 shows a multiple channel fluorogenic response toward fluoride and cyanide and also a selective visual colorimetric response toward cyanide. By comparison with model systems and from detailed photophysical studies on 2 and 3, we conclude that the preferential binding of fluoride occurs at the peripheral borane moieties resulting in the cessation of the EET (electronic energy transfer) process from borane to porphyrin core and with negligible negetive cooperative effects. On the other hand, cyanide binding occurs at the Zn(II) core leading to drastic changes in its absorption behavior which can be followed by the naked eye. Such changes are not observed when the boryl substituent is absent (e.g., Zn-TPP and TPP). Compounds 2 and 3 were also found to be capable of extracting fluoride from aqueous medium.

Conserved water-mediated recognition and dynamics of NAD(+) (carboxamide group) to hIMPDH enzyme: water mimic approach toward the design of isoform-selective inhibitor

Inosine monophosphate dehydrogenase (IMPDH) enzyme involves in GMP biosynthesis pathway. Type I hIMPDH is expressed at lower levels in all cells, whereas type II is especially observed in acute myelogenous leukemia, chronic myelogenous leukemia cancer cells, and 10 ns simulation of the IMP-NAD(+) complex structures (PDB ID. 1B3O and 1JCN) have revealed the presence of a few conserved hydrophilic centers near carboxamide group of NAD(+). Three conserved water molecules (W1, W, and W1 `) in di-nucleotide binding pocket of enzyme have played a significant role in the recognition of carboxamide group (of NAD(+)) to D274 and H93 residues. Based on H-bonding interaction of conserved hydrophilic (water molecular) centers within IMP-NAD(+)-enzyme complexes and their recognition to NAD(+), some covalent modification at carboxamide group of di-nucleotide (NAD(+)) has been made by substituting the -CONH(2)group by -CONHNH2 (carboxyl hydrazide group) using water mimic inhibitor design protocol. The modeled structure of modified ligand may, though, be useful for the development of antileukemic agent or it could be act as better inhibitor for hIMPDH-II.

Gammatone Wavelet Cepstral Coefficients for Robust Speech Recognition

We develop noise robust features using Gammatone wavelets derived from the popular Gammatone functions. These wavelets incorporate the characteristics of human peripheral auditory systems, in particular the spatially-varying frequency response of the basilar membrane. We refer to the new features as Gammatone Wavelet Cepstral Coefficients (GWCC). The procedure involved in extracting GWCC from a speech signal is similar to that of the conventional Mel-Frequency Cepstral Coefficients (MFCC) technique, with the difference being in the type of filterbank used. We replace the conventional mel filterbank in MFCC with a Gammatone wavelet filterbank, which we construct using Gammatone wavelets. We also explore the effect of Gammatone filterbank based features (Gammatone Cepstral Coefficients (GCC)) for robust speech recognition. On AURORA 2 database, a comparison of GWCCs and GCCs with MFCCs shows that Gammatone based features yield a better recognition performance at low SNRs.

RAPID HUMAN ACTION RECOGNITION IN H.264/AVC COMPRESSED DOMAIN FOR VIDEO SURVEILLANCE

This paper discusses a novel high-speed approach for human action recognition in H. 264/AVC compressed domain. The proposed algorithm utilizes cues from quantization parameters and motion vectors extracted from the compressed video sequence for feature extraction and further classification using Support Vector Machines (SVM). The ultimate goal of our work is to portray a much faster algorithm than pixel domain counterparts, with comparable accuracy, utilizing only the sparse information from compressed video. Partial decoding rules out the complexity of full decoding, and minimizes computational load and memory usage, which can effect in reduced hardware utilization and fast recognition results. The proposed approach can handle illumination changes, scale, and appearance variations, and is robust in outdoor as well as indoor testing scenarios. We have tested our method on two benchmark action datasets and achieved more than 85% accuracy. The proposed algorithm classifies actions with speed (>2000 fps) approximately 100 times more than existing state-of-the-art pixel-domain algorithms.

Finding hidden females in a crowd: Mate recognition in fig wasps

Multi-species mating aggregations are crowded environments within which mate recognition must occur. Mating aggregations of fig wasps can consist of thousands of individuals of many species that attain sexual maturity simultaneously and mate in the same microenvironment, i.e, in syntopy, within the close confines of an enclosed globular inflorescence called a syconium - a system that has many signalling constraints such as darkness and crowding. All wasps develop within individual galled flowers. Since mating mostly occurs when females are still confined within their galls,, male wasps have the additional burden of detecting conspecific females that are ``hidden'' behind barriers consisting of gall walls. In Ficus racemosa, we investigated signals used by pollinating fig wasp males to differentiate conspecific females from females of other syntopic fig wasp species. Male Ceratosolen fusciceps could detect conspecific females using cues from galls containing females, empty galls, as well as cues from gall volatiles and gall surface hydrocarbons. In many figs, syconia are pollinated by single foundress wasps, leading to high levels of wasp inbreeding due to sibmating. In F. racemosa, as most syconia contain many foundresses, we expected male pollinators to prefer non-sib females to female siblings to reduce inbreeding. We used galls containing females from non-natal figs as a proxy for non-sibs and those from natal figs as a proxy for sibling females. We found that males preferred galls of female pollinators from natal figs. However, males were undecided when given a choice between galls containing non-pollinator females from natal syconia and pollinator females from non-natal syconia, suggesting olfactory imprinting by the natal syconial environment. (C) 2013 Elsevier Masson SAS. All rights reserved.

FOLD TYPE II PYRIDOXAL 5 `-PHOSPHATE DEPENDENT ENZYMES: STRUCTURE, SUBSTRATE RECOGNITION AND CATALYSIS

Enzymes utilizing pyridoxal 5'-phosphate dependent mechanism for catalysis are observed in all cellular forms of living organisms. PLP-dependent enzymes catalyze a wide variety of reactions involving amino acid substrates and their analogs. Structurally, these ubiquitous enzymes have been classified into four major fold types. We have carried out investigations on the structure and function of fold type I enzymes serine hydroxymethyl transferase and acetylornithine amino transferase, fold type n enzymes catabolic threonine deaminase, D-serine deaminase, D-cysteine desulfhydrase and diaminopropionate ammonia lyase. This review summarizes the major findings of investigations on fold type II enzymes in the context of similar studies on other PLP-dependent enzymes. Fold type II enzymes participate in pathways of both degradation and synthesis of amino acids. Polypeptide folds of these enzymes, features of their active sites, nature of interactions between the cofactor and the polypeptide, oligomeric structure, catalytic activities with various ligands, origin of specificity and plausible regulation of activity are briefly described. Analysis of the available crystal structures of fold type II enzymes revealed five different classes. The dimeric interfaces found in these enzymes vary across the classes and probably have functional significance.

Performance enhancement of online handwritten Tamil symbol recognition with reevaluation techniques

In this article, we aim at reducing the error rate of the online Tamil symbol recognition system by employing multiple experts to reevaluate certain decisions of the primary support vector machine classifier. Motivated by the relatively high percentage of occurrence of base consonants in the script, a reevaluation technique has been proposed to correct any ambiguities arising in the base consonants. Secondly, a dynamic time-warping method is proposed to automatically extract the discriminative regions for each set of confused characters. Class-specific features derived from these regions aid in reducing the degree of confusion. Thirdly, statistics of specific features are proposed for resolving any confusions in vowel modifiers. The reevaluation approaches are tested on two databases (a) the isolated Tamil symbols in the IWFHR test set, and (b) the symbols segmented from a set of 10,000 Tamil words. The recognition rate of the isolated test symbols of the IWFHR database improves by 1.9 %. For the word database, the incorporation of the reevaluation step improves the symbol recognition rate by 3.5 % (from 88.4 to 91.9 %). This, in turn, boosts the word recognition rate by 11.9 % (from 65.0 to 76.9 %). The reduction in the word error rate has been achieved using a generic approach, without the incorporation of language models.