Funded by : Department of Electronics and Information Technology, Government of India
Project Investigator Dr. S Ranbir Singh, Prof. Sukumar Nandi, Dr. Priyankoo Sarmah,
Online demos:
Funded by : Department of Electronics and Information Technology, Government of India
Project Investigator Dr. S Ranbir Singh, Prof. Sukumar Nandi
Project Staffs Ranjan Sarmah, Nitesh Bhattacharjya

Project Description :

We propose to develop a multi-modal broadcast analytics system which can detect and track events of security concern from continuous flow of multimodal data streams originating from hetrogeneous data sources. These multi-modal sources include TV news channels and news websites. Due to multi-modal nature of the data streams our system consists of three sub-systems viz. video analysis sub-system, audio analysis and speech recognition sub-system and Text analytics sub-system

Input to the System :

The input of the broadcast analytics system is the multi-modal data available from various heterogeneous sources such as TV news channels and web news articles. The multi-modal data consists of video, audio, images and textual data. The broadcast news videos are processed beforehand to segment the video streams into news stories and to extract the overlay text, speech transcript and meta-data.

Output of the System : The multi-modal broadcast analytics system has following two types of outputs.
  • Multi-modal event retrieval against free form text query
  • Timeline creation for profiling the entities of interest such as, terror groups, ter- rorist organizations.
Online demos:
Funded By: Department of Electronics and Information Technology, Government of India
Project Investigator : Dr. Sanasam Ranbir Singh
Project Staffs : Rajlakshmi Saikia,Loitongbam Gyanendro Singh, Nanaobi Huidrom, Mayanglambam Bidyalakshmi Devi

Project Description

Speech synthesis is the artificial production of human speech.Text to speech (TTS) convention transforms linguistic information stored as text into speech. Objective of Manipuri TTS synthesis is to convert an arbitrary Manipuri text into corresponding spoken waveform.Text processing and speech generation are two main components of a text to speech synthesis system.In our work we mainly focusing on Concatinative Unit-Selection and Parametric Synthesis system.

For Concatinative Unit-Selection based synthesis we have used openly available Festival framework.Concatenative synthesis is based on the concatenation of units selected from the database.Unit selection synthesis uses large databases of recorded speech.The basic speech segments of our systems are phones (consonants, short vowels and long vowels), bi-phones (CV- clusters: speech segments with a consonant followed by a vowel) and syllables. Indian languages are mostly syllabic in nature, so we have considered syllables as main segmented units.Segmentation is done using Hybrid segmentation(HMM with Group Delay Method).An index of the units in the speech database is then created based on the segmentation and acoustic parameters like the fundamental frequency (pitch), duration, position of syllables in a word(since syllables present in beginning, middle and end of a word has different energy levels).In addition to this, to provide more linguistically balanced clusters, we have also considered textual informations like neighboring syllables of a particular syllable.During synthesis, the desired target utterance is created by determining the best chain of candidate units from the database.Selection of units is done using classification and regression(CART) tree.

Flite is designed as an alternative synthesis engine to Festival.It is primarily designed for small embedded machines and large servers.

In Parametric Synthesis system, we have considered Hidden Markov Model based synthesis(HTS).In the training phase of HTS,spectral parameters and excitation parameters are extracted from the speech data.Using these features and the time-aligned phonetic transcriptions, context-independent monophone HMMs are trained .In the synthesis phase, context-dependent label files are generated for the given text and the required context-dependent HMMs are concatenated.Speech waveform is synthesized directly from the generated spectral and excitation parameters using mel log spectrum approximation(MLSA) filter.

Project Link:Click Here
Project Duration :2012-2015
Funding Agency :MHRD, Department of Higher Education
Project Investigators : Dr. Sanasam Ranbir Singh, Dr. T. Venkatesh
Project Staff :Buddha Saikia,Sisir Kumar Kalita,Gitanjal Bhattacharjya,Burnishwar Nameirakpam, Dirina Gogoi
Project Description :

Aakash project at IIT Guwahati was dedicated to the development of useful applications and contents for use with Aakash. It attempts to empower teachers by using a unique blend of technology, e-contents, and an innovative pedagogy.

Projects developed under the Aakash Application Development Project
  • Examination Conducting System
  • Interactive C
  • Note App
  • eDiscussion
Information explosion and open information sharing by general population over various social media sites pose new challenges to nations in dealing with national security and public safety. One of the core challenges of counter- terrorism and homeland security agencies is to manage the huge stream of information from various information sources in real time, and extracting intelligence to make sense of the information. According to the analysis of B. Raman, a former head of the Counter-terrorism Division of the Research and Analysis Wing (R&AW) in India, Mumbai terror attack has demonstrated the power of effective use of social media and communication media by the terrorist to take advantage for gaining their political agenda. Like Mumbai attackers, organizations like Al-Qaeda, ISIS are using social media to their advantage for various purposes; political propaganda, recruitment, source of critical information, exchange of ideas, raise funds, launder money, share terrorist manuals, and train terrorist members etc. While the above discussion points out, how social and communicative media can be effectively used by terrorist organization to their advantage for executing terrorist activities, several studies have also observed that social and communicating media can also be effectively used to combat terrorism, tracking and prevention. Even though the importance of using Social Network Analysis in counter-terrorism analysis has been recognized much before September 2001 attack, but after 9/11, it has drawn attention from many researchers in academic, government agencies and mainstream media to analyse and understand the structural characteristics of terrorist network. Majority of the datasets used in counter-terrorism studies in the past are homogeneous in nature. However, the nature of information available in social media is heterogeneous consisting of different types of attributes describing the same event. In this study, we proposed a social network analysis framework capable of capturing the heterogeneous characteristics of terrorist attack. Furthermore we predict and analyse relationship between different attributes and organizations constituting a terrorist attack. Some of the relations that we focus in this study are:
  • Chances of a terrorist organisation attacking a particular country or city in future.
  • If a terrorist organisation attacks in future, the chances of attacking a target type such as government building,crowded place etc.
  • Analysis of different terrorist organizations for finding similarities between them.
Further, the proposed method provide a generalisation solution, capable of analysing the network from different perspective without modifying the underlying model. We use Global Terrorist Data (GTD), which has majority of the terrorist attack information between 1970 to 2014.
Online demos:

We often see a political party splitting, government instability, or a politician changing his/her party. In our study, we aim to predict integrity of candidates within a political party using Twitter posts, called tweets, by Indian politicians which in turn would help us predict the above mentioned phenomena. We identify important topics/issues on which the politicians have commented and classify those as either expressing Positive/Negative/Neutral opinion using Sentiment Analysis. We use statistical inference on the results to find how integrated the candidates are. We aim to extend the study to different parties and to journalists.