Virtual Assistant
The current interfaces on a mobile phone are based on alpha-numeric keys and touch technologies, which expects a human being to operate over a tiny area and are extremely suboptimal & restricted for literate/ illiterate and differently-abled. Although, some Virtual Assistants (VAs) have made inroads into the mobile space, but current state-of-the-art VAs suffer from severe limitations including lack of multimodal communication, support for Indian languages and differently-abled individuals. Virtual Assistant project is trying to overcome this limitation and build a rich interface using voice and gesture based technologies for a multi-modal mobile interaction and computing. This project is developing various techniques to augment speech application with dialogue and gesture based interaction in order to improve system performance among patients (users) suffering from various ailments that may impair normal speech. The research work encompasses speech signal processing, acoustic modeling, language modeling, dialog modeling, natural language generation and speech synthesis. This system will act as the baseline for next versions of Virtual Assistant health care system. The overall system will be made to successfully work in real-time for Indian languages even on low-end Android mobile.
MENTORS:
- Prof. Thomas Binford, Chairman and CTO, Read-Ink Technologies Pvt. Ltd.
- Dr. Archan Misra, SMU, Singapore
- Prof. Aloknath De, CTO, Samsung Engineering Pvt. Ltd.
COLLABORATORS:
- Shreshtha Nursing Home
- Vivekanand Hospital iii. Sri City Medical Center
Institutes: IIIT-HYD
Remarks: Avg. enrolment per offering: 60
COURSE TOPICS:
Background and need for speech processing, Speech production mechanism, Nature of speech signal, Basics of digital signal processing, Equivalent representations of signal and systems, Speech signal processing methods, Linear prediction analysis, Basics of speech recognition.
Institutes: IIIT-HYD
Remarks: Avg. enrolment per offering: 50-80
OBJECTIVE :
This is the advanced course in Natural Language Processing intended for honors, dual degree, BTP, MTech and PhD students.
COURSE TOPICS :
In this course, students get an overview of various areas in NLP and the current research trends in each of them. The topics covered include machine translation (rule based & statistical), discourse, statistical parsing, word sense disambiguation, natural language generation, coreference resolution, semantic role labeling etc.. The course also covers two of the most popular machine learning methods (Expectation- Maximization and Maximum Entropy Models) for NLP. Students would be introduced to tools such as NLTK, CoreNLP to aid them in their research.
Institutes: IIIT-HYD
Remarks: Avg. enrolment per offering: 15-20
Objective:
This is an advance course whose objective is to discuss and provide hands-on experience on implementation of algorithms, models used in feature extraction and in building speech systems.
COURSE TOPICS:
1.Introduction to speech technology
2.Feature extraction from speech signal
3.Algorithms for speech recognition
4.Methods for speech synthesis
5.Approaches for speech enhancement
6.Approaches for speaker recognition
Institutes: IIIT-HYD
Remarks: Avg. enrolment per offering: 40-60
OBJECTIVE :
This is the advanced course in Natural Language Processing intended for honors, dual degree, BTP, MTech and PhD students.
COURSE TOPICS :
In this course, students get an overview of various areas in NLP and the current research trends in each of them. The topics covered include machine translation (rule based & statistical), discourse, statistical parsing, word sense disambiguation, natural language generation, coreference resolution, semantic role labeling etc.. The course also covers two of the most popular machine learning methods (Expectation-Maximization and Maximum Entropy Models) for NLP. Students would be introduced to tools such as NLTK, CoreNLP to aid them in their research.
Institutes: IIIT-HYD
Remarks: Avg. enrolment per offering: 40-60
OBJECTIVE:
This course is intended to be an advanced course for graduate and senior MS students planning to pursue research in the areas of NLP and IR. The course draws heavily from the latest research papers and tutorials covering the state-of-the-art techniques and problems in these areas.
COURSE TOPICS:
1.Deep Learning for NLP
a.RNN based Language models
b.LSTM for Language Modelling
c.Deep NNs for POS tagging, Chunking and NER
2.Deep Structured Semantic Models
a.Deep Learning for Semantic Matching
b.Learning Deep structured models for Search
c.CNNs for DSSMs
3.Multilingual IR
a.Cross Lingual IR
b.Multilingual PRF
c.Multilingual Topic Modelling
4.Learning to Rank
a.Pointwise, Pairwise and Listwise Ranking Paradigms
b.Discriminative Models (SVMs, MeMM, RankSVM, IRSVM
) c.Ranking Evaluation- Online & Offline Techniques
5.Statistical Translation Models for IR
a.IR as SMT
b.Word based Translation Models (TMs)
c.Phrase-based TMs, Syntax-based TMs
Institutes: VNRVJIET-HYD
Remarks: Course Objectives
To introduce speech production and related parameters of speech.
To show the computation and use of techniques such as short time Fourier transform, linear predictive coefficients and other coefficients in the analysis of speech.
ï”o understand different speech modeling procedures such as Markov and their implementation issues.
Title: INTERNET OF THINGS
Institutes: VNRVJIET-HYD
Remarks: Course Objectives
To understand the basics of Internet of Things
To get an idea of some of the application areas where Internet of Things can be applied
To understand the concepts of web and middleware for Internet of Things
To understand the concepts of Cloud of Things with emphasis on Mobile cloud computing and IOT protocols
Institutes: IIIT-HYD
Remarks: Objectives
- Embedded Linux platforms, IOT applications and related case studies. Interfacing various Digital/Analog Sensors (GPIO, PWM, Analog). Understanding Intel Galileo and its Interfacing techniques. Embedded protocols and setting up IoT components: As part of project case study students performed Data logging using SD library and Programming UART based GPS/GSM-GPRS Shields and related applications.
- Building IoT SD card image, Accessing Linux terminal using Virtual Serial Interface like PUTTY/Tera term (Shell commands and utilities) and Setting up Networking (*Ethernet or *Wi-Fi). Getting started with python programming on Galileo, UPM and MRAA for Intel Galileo Gen 2 and Application building using python.
- Project Ideation and Building IoT based Applications Building POSIX based Embedded Linux applications, 3-Tier architecture of IoT applications and related protocols: Understanding TCP/IP and Arduino Ethernet Library API and HTTP based Arduino applications.
- Building IoT applications: ThingSpeak cloud management framework for data logging and data visualization and Dweet based IoT application building. Twitter integration to control devices using Intel Galileo Gen 2 board and GMAIL server Integration with Intel Galileo Building Prototypes using Intel Galileo. Evaluation and assessment activities, Project idea demonstration and Feedback and conclusion
Institutes: IIIT-HYD
Remarks: Avg. enrolment per offering: 90-110
In this course, students get an overview of various areas in NLP and the current research trends in each of them. The topics covered include machine translation (rule based & statistical), discourse, statistical parsing, word sense disambiguation, natural language generation, coreference resolution, semantic role labeling etc.. The course also covers two of the most popular machine learning methods (Expectation- Maximization and Maximum Entropy Models) for NLP. Students would be introduced to tools such as NLTK, CoreNLP to aid them in their research. Popular deep learning models for NLP are also introduced during this course.
Institutes: IIIT-HYD
Remarks: Avg. enrolment per offering: 40-60
This course is intended to be an advanced course for graduate and senior MS students planning to pursue research in the areas of NLP and IR. The course draws heavily from the latest research papers and tutorials covering the state-of-the-art techniques and problems in these areas.
COURSE TOPICS:
1.Deep Learning for NLP
2.Deep Structured Semantic Models
3.Multilingual IR
4.Learning to Rank
5.Statistical Translation Models for IR
Institutes: VNRVJIET-HYD
Remarks: Registration date: 24-08-2017
Institutes: VNRVJIET-HYD
Remarks: Application Number: 201741045066 Filing
Date: December 14, 2017
Remarks: Date: 17-20 January
Participants: 150
Remarks: Date: 13 December
Participants: 150
Remarks: Date: July 1st - July 15th,2014
Participants: 80
Remarks: Date: 12 December
Participants: 150
Remarks: Date: 1-15 July 2015
Participants: 87
Remarks: Date: 24 December
Participants: 150
Remarks: Date: 20th June to 4th July, 2016
Participants: 80
Remarks: Date: 24 May-8 June 2017
Participants: 109
Remarks: Date : 17th December 2014
Participants: 50
Links/URL: http://www.ide.iitkgp.ernet.in/Pedagogy1/course_details.jsp
Title: Basic Electronics
Links/URL: http://www.ide.iitkgp.ernet.in/Pedagogy1/course_details.jsp
Remarks: Date: 13 th 16 th December 2016
Participants: 50
Remarks: Date: 29 th - 31 st October 2015
Title: Recent Trends in Signal Processing and Its applications Remarks: Date: 20-06-2017 to 30-06-2017
Participants: 48
Remarks: Date: (i) 29-05-2017 to 06-06-2017 (ii) 07-06-2017 to 14-06-2017
Participants: (i)72 (ii) 80
Institutes: IIIT-HYD
Remarks: Year: 2016
Title: IIIT-H R&D Showcase 2015/2016/2017/2018Institutes: IIIT-HYD
Remarks: Year: 2015, 2016, 2017, 2018
Title: National workshop on ICT initiatives for Rural Development by NIRD and Media Lab AsiaInstitutes: VNRVJIET-HYD
Remarks: Date: Sep 1, 2017
Title: International Innovation Fair AssociationInstitutes: VNRVJIET-HYD
Remarks: Held at Vishakhapatnam during 9-11th Sep 2017
Title: NASSCOM Product ConclaveInstitutes: VNRVJIET-HYD
Remarks: During Oct 26th- 27th, 2016 at Bangalore
Title: AP & TS sections 12 th Annual Convention and Awards Function-2017Institutes: VNRVJIET-HYD
Remarks: Held on 29th December, 2017 at AITAM, Tekkali
Title: 2nd World Invention and Innovation ForumInstitutes: VNRVJIET-HYD
Remarks: Date: November 23-25, 2017
Title: Seoul International Invention Fair (SIIF)Institutes: VNRVJIET-HYD
Remarks: Date: 1-4th December 2016.
Remarks: Collaboration Area: Knowledge graph based Virtual Assistant.
Remarks: Collaboration Area: We received support for the Cough Analysis work from Azim Premji Foundation.
Remarks: Offered to: FlyDubai
Developed a Goal oriented chatbot for FlyDubai Airlines to be deployed at their premises. The bot was an extension of the technology developed during the course of the project. The engine and chat framework was re-engineered by our students as prototype for FlyDubai.
Remarks: Offered to : TalkADoc Inc.
A startup was incubated to develop a chatbot to track mental health status for perinatal care. This bot is designed to allow doctors to have a way to oversee the progress of the patients without direct contact. The chatbot is non-transactional, information gathering agent which interacts with the patient over an long duration providing automatic summaries to the doctor.
Remarks: Collaboration Area: Testing and prototype development of cough analysis module
Title: Sai Siddarth hospitalsRemarks: Collaboration Area: Data samples collection for cough analysis
Manish Srivastava(IIIT-HYD), Principal Investigator
Anoop Namboodri(IIIT-HYD), ITRA Faculty | 1. | Significance of neural phonotactic models for large-scale spoken language identification |
| 2. | Analysis of Cough Sounds for Diagnosis of Respiratory Diseases |
| 3. | Early Diagnosis of Pulmonary Diseases using Wheeze Sound Analysis |
| 4. | Expert system using artificial neural network for chronic respiratory diseases |
| 5. | Together we stand: Siamese networks for similar question retrieval |
| 6. | Articulatory gesture rich representation learning of phonological units in low resource settings |
| 7. | Mirror on the wall: Finding similar questions with deep structured topic modeling |