International Institute of Information Technology, Hyderabad

About: International Institute of Information Technology, Hyderabad is a education organization based out in Hyderabad, India. It is known for research contribution in the topics: Authentication & Internet security. The organization has 2048 authors who have published 3677 publications receiving 45319 citations. The organization is also known as: IIIT Hyderabad & International Institute of Information Technology (IIIT).

Poster: LWE: LDA Refined Word Embeddings for Duplicate Bug Report Detection

Abstract: Bug reporting is a major part of software maintenance and due to its inherently asynchronous nature, duplicate bug reporting has become fairly common. Detecting duplicate bug reports is an important task in order to avoid the assignment of a same bug to different developers. Earlier approaches have improved duplicate bug report detection by using the notions of word embeddings, topic models and other machine learning approaches. In this poster, we attempt to combine Latent Dirichlet Allocation (LDA) and word embeddings to leverage the strengths of both approaches for this task. As a first step towards this idea, we present initial analysis and an approach which is able to outperform both word embeddings and LDA for this task. We validate our hypothesis on a real world dataset of Firefox project and show that there is potential in combining both LDA and word embeddings for duplicate bug report detection.

MARASIM: a novel jigsaw based authentication scheme using tagging

Abstract: In this paper we propose and evaluate Marasim, a novel Jigsaw based graphical authentication mechanism using tagging. Marasim is aimed at achieving the security of random images with the memorability of personal images. Our scheme relies on the human ability to remember a personal image and later recognize the alternate visual representations (images) of the concepts occurred in the image. These concepts are retrieved from the tags assigned to the image. We illustrate how a Jigsaw based approach helps to create a portfolio of system-chosen random images to be used for authentication. The paper describes the complete design of Marasim along with the empirical studies of Marasim that provide evidences of increased memorability. Results show that 93% of all participants succeeded in the authentication tests using Marasim after three months while 71% succeeded in authentication tests using Marasim after nine months. Our findings indicate that Marasim has potential applications, especially where text input is hard (e.g., PDAs or ATMs), or in situations where passwords are infrequently used (e.g., web site passwords).

Why does substitution of thymine by 6-ethynylpyridone increase the thermostability of DNA double helices?

Abstract: Efficiency of 6-ethynylpyridone (E), a potential thymine (T) analogue, which forms high-fidelity base pairs with adenine (A) and gives rise to stabler DNA duplexes, with stability comparable to those containing canonical cytosine(C):guanine(G) base pairs, has been reported recently. Estimates of the interaction energies, involving geometry optimization at the DFT level (including middle range dispersion interactions) followed by single point energy calculation at MP2 level, in excellent correlation with the experimentally observed trends, show that E binds more strongly and more discriminately with A than T does. Detailed analysis reveals that the increase in base–base interaction arises out of conjugation of acetylenic π electrons with the ring π system of E, which results in not only an extra stabilizing C–H···π interaction in the EA pair, but also a strengthening of the conventional hydrogen bonds. However, the computed base–base interaction energy for the EA pair was found to be much less than that of...

Investigative study of various activation functions for speech recognition

Abstract: Significant developments in deep learning methods have been achieved with the capability to train more deeper networks. The performance of speech recognition system has been greatly improved by the use of deep learning techniques. Most of the developments in deep learning are associated with the development of new activation functions and the corresponding initializations. The development of Rectified linear units (ReLU) has revolutionized the use of supervised deep learning methods for speech recognition. Recently there has been a great deal of research interest in the development of activation functions Leaky-ReLU (LReLU), Parametric-ReLU (PReLU), Exponential Linear units (ELU) and Parametric-ELU (PELU). This work is aimed at studying the influence of various activation functions on speech recognition system. In this work, a hidden Markov model-Deep neural network (HMM-DNN) based speech recognition is used, where deep neural networks with different activation functions have been employed to obtain the emission probabilities of hidden Markov model. In this work, two datasets i.e., TIMIT and WSJ are employed to study the behavior of various speech recognition systems with different sized datasets. During the study, it is observed that the performance of ReLU-networks is superior compared to the other networks for the smaller sized dataset (i.e., TIMIT dataset). For the datasets of sufficiently larger size (i.e., WSJ) performance of ELU-networks is superior to the other networks.

Learning speaker-specific phrase breaks for text-to-speech systems.

Abstract: The objective of this paper is to investigate whether prosodic phrase breaks are specific to a speaker, and if so, propose a mechanism of learning speaker-specific phrase breaks from the speech database. Another equally important aspect dealt in this work is to demonstrate the usefulness of these speaker-specific phrase breaks for a text-to-speech system. Experiments are carried out on two different English voices as well as on a Telugu voice, and it is shown that speaker-specific phrase breaks improves duration as well as spectral quality of synthetic speech.