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).

Architectural space planning using Genetic Algorithms

Abstract: This paper presents a system which can find out space planning for a single flat, arrangement of several flats on a single floor and extend the design for each floor and find out collective plan for a multi-storey apartment building. At each level it generates a plan which supports quick evacuation in case of adversity. Starting with design specifications in terms of constraints over spaces, use of Genetic Algorithm leads to a complete set of consistent conceptual design solutions named topological solutions. These topological solutions which do not presume any precise definitive dimension correspond to the sketching step that an architect carries out from the design specifications on a preliminary design phase in architecture. Further, door placement algorithm has been proposed with modifications in existing Dijkstra's algorithm and dimensions analysis is carried out for the designs selected by the user. If the user wishes to generate a plan for many floors, inputs are taken accordingly and plan is generated which is efficient in terms of evacuation.

Providing Web Credibility Assessment Support

Abstract: Presence of information from multiple sources on the internet requires evaluating the credibility of the information, before its utilization. Researchers have suggested that internet users experience difficulty in accessing necessary information and do not pay enough attention to its credibility. We present here the design and implementation of an automated Web Credibility Assessment Support Tool (WebCAST) that considers multiple factors (type of website, popularity, sentiment, date of last update, reputation and review based on users' ratings reflecting personal experience) for assessing the credibility of information and returns a summary indication of the credibility of a website. We use Potentially All Pairwise RanKings of all possible Alternatives (PAPRIKA) method of Multi-Criteria Decision Analysis (MCDA) to give weights to the scale values on each factor, representing the relative importance of the attributes. An empirical evaluation of the tool was conducted by computing the correlation between the tool-generated credibility scores and that of human judges. The correlation was found to be 0.89, thus verifying the validity of the tool. In the future the proposed tool can be made useful to students in their learning process of credibility assessment.

Multimodal Sentiment Analysis Using Deep Neural Networks

Abstract: Due to increase of online product reviews posted daily through various modalities such as video, audio and text, sentimental analysis has gained huge attention. Recent developments in web technologies have also enabled the increase of web content in Hindi. In this paper, an approach to detect the sentiment of an online Hindi product reviews based on its multi-modality natures (audio and text) is presented. For each audio input, Mel Frequency Cepstral Coefficients (MFCC) features are extracted. These features are used to develop a sentiment models using Gaussian Mixture Models (GMM) and Deep Neural Network (DNN) classifiers. From results, it is observed that DNN classifier gives better results compare to GMM. Further textual features are extracted from the transcript of the audio input by using Doc2vec vectors. Support Vector Machine (SVM) classifier is used to develop a sentiment model using these textual features. From experimental results it is observed that combining both the audio and text features results in improvement in the performance for detecting the sentiment of an online product reviews.

Person de-identification in videos

Abstract: Advances in cameras and web technology have made it easy to capture and share large amounts of video data over to a large number of people through services like Google Street View, EveryScape, etc A large number of cameras oversee public and semi-public spaces today These raise concerns on the unintentional and unwarranted invasion of the privacy of individuals caught in the videos To address these concerns, automated methods to de-identify individuals in these videos are necessary De-identification does not aim at destroying all information involving the individuals Its goals are to obscure the identity of the actor without obscuring the action This paper outlines the scenarios in which de-identification is required and the issues brought out by those We also present a preliminary approach to de-identify individuals from videos A bounding box around each individual present in a video is tracked through the video An outline of the individuals is approximated by carrying out segmentation on a 3-D Graph of space-time voxels We explore two de-identification transformations: exponential space-time blur and line integral convolution We show results on a number of public videos and videos collected in a plausible setting We also present the preliminary results of a user-study to validate the effectiveness of the de-identification schemes.

Hindi to English Machine Translation: Using Effective Selection in Multi-Model SMT

Abstract: Recent studies in machine translation support the fact that multi-model systems perform better than the individual models. In this paper, we describe a Hindi to English statistical machine translation system and improve over the baseline using multiple translation models. We have considered phrase based as well as hierarchical models and enhanced over both these baselines using a regression model. The system is trained over textual as well as syntactic features extracted from source and target of the aforementioned translations. Our system shows significant improvement over the baseline systems for both automatic as well as human evaluations. The proposed methodology is quite generic and easily be extended to other language pairs as well.