Participatory sensing based Android app for increasing cultural harmony

Extending the work of developing a mobile application for increasing cultural interaction by using augmented reality. This will be an extension of the current SanskAR app, by including participatory sensing to enhance the knowledge creation and dissemination. The student will have to work on the technical aspect as well as social aspect of the project. A cross-cultural experience would be advantageous for the applicants.

Design & Implementation of an affordable ultrasonic navigation system for the Blind

The goal of the project is to explore different ultrasonic sensor designs, design, interface with a white cane and develop an indoor and outdoor navigation system for the blind. The main aim is to make the unit affordable as those in the market at highly priced leaving the blind and visually impaired to resort to just a white cane.

Real-time classification of equipments in a smart building

This research work aims in implementing a real-time equipment classification algorithm for smart buildings. This algorithm will be later integrated to a context aware system for smart building to achieve  profit maximization in smart buildings. This work involved real-deployment of the system and experimentation.

Analysis and implementation of vertical handoff mechanism for large scale monitoring application

In this research work initially study and analysis of already existing vertical handoff mechanism is performed. After that the best vertical handoff mechanism is chosen and enhanced with respect to the requirement of the large scale monitoring application. The new mechanism is implemented and evaluated

Self-Organizing Wireless Sensor Network for Monitoring Natural Disasters

Disasters are ever increasing in an alarming rate. Amrita University has deployed a wireless sensor network for monitoring and detection of landslide in India. This network may experiences link failures due to several factors such as environmental conditions, non availability of power, hardware failures etc. This research work's objective is to design, and implement an innovative self-organizing protocol for large scale wireless networks. These networks are constrained by the energy availability and they have to work in near real-time. The protocol should cater to these scenarios.

Seamless Connectivity in Heterogeneous Wireless Networks

Heterogeneous Networks are required for long distance transmission of data from large scale wireless sensor network for monitoring disasters. In the research project, the objective is to design and implement a seamlessly connected heterogeneous wireless network to better disaster management and communications. Implementation and performance evaluation will be part of this work.

Context Aware Cypher Physical System for Sustainable Smart Building

In this research work a context aware cypher physical system has to be designed and implemented. This system should be capable to achieve energy sustainability in smart building.  Sensor networks and algorithms has to be developed and implemented for achieving this goal.

Open innovation and the innovation performance of SMEs in the food product manufacturing industries

There have been a number of studies of SME innovation suggesting that such firms see innovation differently from larger firms and governments. The overarching research question for the proposed study is: what are the sources of innovation for SMEs in the food products manufacturing industry? It is intended this study will complement those associated with automotive component and other similar manufacturers.

Five research objectives have been formulated to address the key dimensions of this research question:

  1. To investigate the level and nature of innovation activity, its associated discourse and resulting innovation performance within SMEs.
  2. To identify and map knowledge inflows and outflows associated with innovation activity within SMEs.
  3. To investigate the capacity of SMEs to access, evaluate, assimilate and exploit new knowledge in order to facilitate and effect innovation activity and performance.
  4. To explore the influence of local conditions (the business and policy context) on innovation activity and performance within SMEs.
  5. To identify ‘open innovation accelerators’ which facilitate and enhance knowledge flows and innovation activity within SMEs.
     

VLSI Testing of SoC designs using Logic BIST

Logic Built in Self Test (LBIST) is the best alternate to the conventional Automatic Test Equipment (ATE) based testing of Electronic circuits. LBIST brings the advantages of reduction in test time, reduction in test cost, at speed testing (ability to apply a huge number of test patterns), on site testing, reduced storage and maintenance comparing to ATE based testing. Logic BIST is relatively a new topic in the domain of Testing of Electronic Circuits.

This project is aimed to design a reconfigurable logic built in self-test technique allowing the circuit to perform self-testing i.e., testing of their own operation (functionally, parametrically, or both) using their own circuits. Features such as boundary scan and bit fixing logic need to be added to make the LBIST meet with industry standards. Hardware security of the SoC can also be enhanced in this technique.  The work will be carried out using Cadence tools.  The major objectives are

  1. Design reconfigurable Logic BIST blocks and test it for different designs
  2. Secure Boundary scan and Weighted bit fixing logic development
  3. Check the flow through industry standard Cadence Encounter Test Architect.
     

Analysis of emotions in speech through throat microphone and electro-glottogram data

The purpose of the project is to analyze the variations in the throat microphone recorded speech data and electro-glottogram (EGG) data according to different simulated emotion conditions. Also the possibility of feature mapping of throat microphone recorded speech and EGG recorded data with the close speaking micro phone data has to be explored as a part of the project.

Role and Impact of Alumni in a B-School

The Project would try to assess the impact that the B-School alumni/ae have made so far, in the past 17 years of its existence. It would also try to understand how the alumni/ae could help to advance the school's reputation.

A Comparison of Sustainable Development Practices in India and the country from which the Intern comes.

  • This is a very broad subject area, and can host a number of interns, each one focusing on a distinct dimension like water, energy, cities, agriculture etc.
  • The study would be based primarily on secondary data sources, with the possibility of one real case study in each location.
  • The objective of the study would be to learn for the experiences of the two countries, and offer viable alternative solutions.
     

A comparison of Supply Chain Management (SCM) practices in India and the country from which the Intern comes.

The study would look into two similar manufacturing organizations in each country and would try to compare their SCM practices. The objective of the study would be to learn for the experiences of the two cases, and offer a few recommendations for improvements.

The following internship/projects are available currently: A) Electrodeposition of conducting polymers B) Electrochemical synthesis of graphene oxide C) Development of standard voltammetric sensing protocols

All the internships/projects listed below form part of a larger goal of our lab - electrochemical fabrication of non-enzymatic sensing platforms.

A) Electrodeposition of conducting polymers
Electrodeposited conducting polymers form the backbone of electrochemically fabricated sensing platforms. To this end, we are interested in examining electrodeposition of conducting polyaniline (PANi), polypyrrole (PPy), and polyethylenedioxythiophene (PEDOT).

B) Electrochemical synthesis of graphene oxide
Reduced graphene oxide with a well-designed functional group distribution is one of the best non-enzymatic sensing elements as has been repeatedly borne out in the literature, due to its high surface area, excellent sensitivity and even selectivity. The objective of this project/internship is to develop electrochemically reduced graphene oxide.

C) Development of standard voltammetric sensing protocols:
Voltammetric sensing is one of the attractive techniques due to the precise control achievable in both the perturbing signal (voltage) and measured response (current). Voltammetric techniques have been used for detecting trace concentrations (even less than nanomolar levels) of many analytes. The success of a voltammetric sensing protocol depends on the rigor with which uncontrolled variables are eliminated, and errors minimized. Many standard protocols exist - each lab usually has a variation of these protocols.  This work aims at bolstering existing protocols in our lab by an end-to-end analysis of errors.

Interns/students are expected to self-motivated and systematic in their work. They are expected to maintain systematic record of their work and will make periodic presentations to the supervisor.

These projects are also open for collaborative research with students pursuing their PhD in other universities, provided a collaboration is established with the research supervisor/lab they are associated with.

A Hybrid GA Approach to solve Traveling Salesman problem.

  • You are required to study and develop various Genetic Algorithms.
  • Develop Model for the TSP.
  • Experiment Results & Comparison
     

Finite element simulation of Machining processes

This project is aimed at stimulating machining processes such as turning, milling, drilling. Simulation results have to be compared with experimental results. The comparison and successful modeling of the different processes will help us to understand the  complex process behind machining

"Determine Performance Analysis of propagation models in different terrain conditions with MIMO channel for IEEE Standard 802.16e WiMAX"

The spatial channel model is the one that includes multiple antennas profile in the WiMAX system is the 2×2 MIMO technique. The system performs spatial multiplexing by allowing any type of antenna configuration, spatial channel model uses both link-level and system-level simulation for up-link and down-link operation. Spatial channel model takes the multiple-input multiple-output radio link parameters, model configuration parameters, and antenna parameters as inputs, and outputs the MIMO channel matrix. Channel matrix can be generated for multiple Base station-Mobile station. WiMAX has potential success and is gaining hike all over the world for deployment of the network in its line-of-sight (LOS) and non-line-of-sight (NLOS) conditions for both fixed operation below 11GHz and Mobile operation below 6 GHz frequency range. In different topography (urban, suburban, and rural) the model can be implemented by generating the matrix for different channels for transmitter and receiver antennas. Using the MIMO channels matrix are generated for the urban canyon and urban macro area, from this path loss, delay, fading in the channel can be calculated. Channel matrix uses some of the real environmental parameters for the simulation in this work, at the transmitter and receiver system different types of multiple antennas can be used at the specific distance of separation. Spatial channel model generates any dimensional matrix for the reference of antennas and that can be used for the network planning and antenna separation in the channels.

Assessing the Perceptual Behavior of a Person using Eye Gaze Tracker

“People have a model of the world by moving their eyes”. To study the complexities of human behavior requires data collection methods that are discrete as well as accurate. Eye and gaze tracking solutions can assist in monitoring on-screen eye movements in finding objects, and saccades/anti-saccades for both the qualitative and quantitative observation of the perceptual behavioral of a person. The human eye is constantly responding to stimuli and this information can be used for guiding various systems much faster than with conventional input devices. Since visual interaction devices are extremely small, lightweight and mobile, the possibilities for their use and areas of gaze-based interaction research are manifold, and continuously expanding. This work focus on developing an open source eye gaze tracker which assess the visual perception behavior of a person from the eye gaze patterns. The proposed system extends its applications in the area of assistive systems.

Blood Vessel Segmentation from Color Retinal Images using Unsupervised Texture Classification

Automated blood vessel segmentation is an important issue for assessing retinal abnormalities and diagnoses of many diseases. The segmentation of vessels is complicated by huge variations in local contrast, particularly in case of the minor vessels. In this paper, we propose a new method of texture based vessel segmentation to overcome this problem. We use Gaussian and L*a*b* perceptually uniform color spaces with original RGB for texture feature extraction on retinal images. A bank of Gabor energy filters are used to analyze the texture features from which a feature vector is constructed for each pixel. The fuzzy C-means (FCM) clustering algorithm is used to classify the feature vectors into vessel or non-vessel based on the texture properties. From the FCM clustering output we are expecting to attain the final output segmented image after a post processing step. We will compare our method with hand-labeled ground truth segmentation of some images and achieve approx 84.37% sensitivity and 99.61% specificity.

Wavelet-Regularized Reconstruction of Dynamic PET

Tomographic reconstruction from positron emission tomography (PET) data is an ill-posed problem that requires regularization. An attractive approach is to impose an ℓ1-regularization constraint, which favors sparse solutions in the wavelet domain. This can be achieved quite efficiently thanks to the iterative algorithm developed by Daubechies et al., 2004. We apply this technique and extend it for the reconstruction of dynamic (spatio-temporal) PET data. Moreover, instead of using classical wavelets in the temporal dimension, we introduce exponential-spline wavelets (E-spline wavelets) that are specially tailored to model time activity curves (TACs) in PET. We show that the exponential-spline wavelets naturally arise from the compartmental description of the dynamics of the tracer distribution. We demonstrate the usefulness of spatio-temporal regularization and the superior performance of E-spline wavelets over conventional Battle-Lemarié wavelets in a series of experiments: the 1-D fitting of TACs, and the tomographic reconstruction of both simulated and clinical data. The figure shows a spatial and temporal cutting plane with from left to right: conventional filtered back projection, temporal wavelet-regularized reconstruction, and spatio-temporal wavelet-regularized reconstruction.

Data Mining and Pattern Recognition for Large-Scale Scientific Data

Data mining techniques can be applied to data gathered from simulations, experiments, or observations in various scientific domains, The use of automated techniques can help bring objectivity to such data analysis. In addition, data that was originally obtained for one purpose can now be analyzed using pattern recognition techniques to detect previously unknown patterns in the data.

Large-scale scientific data mining is a very challenging field, making it a source of several open research problems. In order to extend PR techniques to large-scale data, several barriers must be overcome. The extraction of key features from large, multi-dimensional, complex data is a critical issue that must be addressed first, prior to the application of the pattern recognition algorithms. The features extracted must be relevant to the problem, insensitive to small changes in the data, and invariant to scaling, rotation, and translation.  In addition, we need to select discriminating features through appropriate dimension reduction techniques. The pattern recognition step poses several challenges as well. For example, is it possible to modify existing algorithms, or design new ones, that are scalable, robust, accurate, and interpretable? Further, can these algorithms be applied effectively and efficiently to complex, multi-dimensional data? 

A capability for large-scale pattern recognition will  provide  an effective way to cope with data overload scaling data mining to large and complex data sets. Pattern Recognition Algorithms for Data Mining addresses different pattern recognition (PR) tasks which  include data condensation, feature selection, case generation, clustering/classification, and rule generation and evaluation. Project can be taken in finding algorithms capable of processing input data related to a specific application.

A Computer-Aided Diagnosis System for Breast Cancer Detection by using a Wavelet and Curvelet techniques

The most common type of cancer among women worldwide is breast cancer. Early detection of breast cancer is very important to reduce the fatality rate. For the hundreds of mammographic images scanned by a radiologist, only a few are cancerous. While detecting abnormalities, some of them may be missed, as the detection of suspicious and abnormal images is a recurrent mission that causes fatigue and eyestrain. In this paper, a computer-aided diagnosis system using the curvelet transform (CT) algorithm is proposed for interpreting mammograms to improve the decision making. The purpose of this study is to develop a method for the characterization of the mammography as both normal and abnormal regions, and to determine its diagnostic performance to differentiate between malignant and benign ones. This work will also include  comparison of wavelet and curvelet for breast cancer diagnosis in digital mammogram

Micro calcifications are tiny deposits of calcium that appear as small bright spots in the mammogram. Although they have higher inherent attenuation properties, they cannot be distinguished from the high-frequency noise because of their small size. So, the relevant features involved are variable, occur at different scales and orientations and are characterized by discontinuous changes in intensity as well as more subtle global variations in texture. For this reason, the first step of the cancer signs detection should be a segmentation procedure able to distinguish masses and micro calcifications from background tissue and noise. This can be done in a wavelet domain, owing to multiscale nature of cancer signs. In fact, using wavelet transform, it is possible to detect details that appear at different scales and selectively enhance them within different resolution levels. Moreover, mass segmentation has to be performed by firstly applying suitable operators, since, because of low contrast, they appear embedded in and camouflaged by varying densities of parenchimal tissue structures. So, it is very difficult not only to visually detect them and their boundaries on mammograms but also to distinguish, in wavelet domain, the coefficients related to suspicious masses from the ones related to background. The diagnostic model will be designed and developed  for the principal component analysis and linear discriminant analysis for further dimensional reduction and feature selection. A dataset from the Mammographic Image Analysis Society database will be  used for testing the method

Automatic Identification of White Stem Borer Infected Coffee Plant.

Walking of a Biped Robot on normal surface has been investigated since two decades. Walking on uneven terrain poses difficult decision and stability concerns for a biped robot. Incorporating sensors for mapping the terrain is still in the early stages of development. Developing a stable walking strategy for a biped robot by combining the terrain information and control takes humanoid robotic research to a new level. Also it has immense practical applications in real world.

The research findings can be used in real world applications in the field of defence, service industries and in hazardous environments for replacing the human workforce.

The objective of the project is to develop, design and implement a real-time feedback control system for a biped robot for a stable and smooth movement over any given terrain including uneven surfaces. This project aims to

  1. Generate a stable reference trajectory which the biped can follow
  2. Develop a feedback controller for tracking the above reference trajectory and
  3. Finally to implement the controller and sensor on a biped robot enabling it to move smoothly over any uneven surface in real time.
     

The preliminary work involves the development of the controller for tracking a stable reference trajectory generated off-line using software simulation studies. The trajectory will be generated after considering the stability constraints using Zero Moment Point analysis. Once the controller is successfully designed, the mapping of terrain information using vision sensors and online calculation of the stable trajectory will be taken up. The machine vision module and the controller module will be integrated together to get a complete model which can help a biped robot to navigate smoothly over any uneven terrain.

Automation of Density Measurement Process

Design and developing a robotic ARM which will pick and place objects of weight ranging from 2mg to 200g in a research lab. Objects is of irregular shapes. The workspace of the robot has several constraints. The robot has to be very precise and accurate in its movements.

Study and Performance Analysis of New Transmission Technique of Analog Communication with Feedback

Analogue signal transmission is considered as forgotten method of communication system (CS) especially when established digital communications is widely accepted and used almost everywhere.  However, neither C.E. Shannon nor the founders of information and communication theory considered digital transmission as more prospective than the analog one. Moreover, in the 1950-60s, the analog communication systems with feedback channels were a subject of very intensive research.  But advancement and impressive success of digital communication technique and applications with plenty of merits not only switched the researchers attention to digital communication systems but also virtually put a full stop to analog counterpart. The success of digital communication technique and systems is visible, however, the system has become more complex both in terms of computational and hardware/software points of view. Digital transmission technique based system is reaching to performance limit level. This has already been debated. Therefore,  the revival of old technology that is Analog communication with feedback or precisely called analog feedback communication systems (AFCS) is very likely to find researcher attention in very near future. In fact, few researchers have already found many relative enhancement in performance parameters as compared to digital counterpart. Where digital transmission technique performance has reached to optimal level and quite impossible without complexity and degradation in power-bandwidth efficiency, the analog transmission technique enjoys the efficiency at the same time on reduced complexity. It is expected that AFCS will be the future of the communication technique.

In this internship program, the student will be doing analytical/mathematical analysis of this new transmission technique, develop the channel model and compare the performance in respect of bit error rate, signal-to-noise ratio requirements with that of existing digital transmission technique

Design of Visible Light Communication System for Video Transmission from Table Lamp

Visible light communication (VLC) is a new emerging technology which uses visible light as a medium to transmit data wirelessly. It is a novel kind of optical wireless system. VLC exploits the advancement in light emitting diode (LED) which has undergone tremendous growth recently. LED-based lighting is being projected as the future lighting technology because of several advantages such as long life, human friendly, energy efficient and so on. Being a semiconductor device, LED has inherent characteristics of high speed switching. VLC leverages these characteristics and make dual usage of LED; lighting and data communication simultaneously. VLC can be designed at low cost and on existing infrastructure or with slight modification on it.

Recently, researches in VLC system have increased significantly. They have several potential applications both indoor and outdoor, yet to be explored. Although, VLC has many novel and unique characteristics, the design is challenging. In its original and most basic form, VLC is suitable for a broadcast system however, bidirectional link design, high data rate, video transmission, long distance communications, are difficult. Though, VLC theoretical offers unlimited bandwidth (frequency range 380-790THz), however, actual rate of transmission is limited by electronics that is switching speed of LED used as VLC transmitter and photo diode and front end amplifier used as a part of VLC receiver.

In this work, we will investigate VLC system for high data rate transmission such as transmission of video. A set of high brightness LED can be integrated as a table lamp which will be a part of VLC transmitter. Similarly, photo diode based VLC receiver will be designed to support such a bandwidth and high data rate transmission. Student(s) will be required to model analytically the optical wireless channel, characterized the VLC emitter and receiver and then hardware circuit design with any electronic device tool such as PSpice, CADENCE etc..Finally, it will be demonstrated to work for video transmission.

Characterization of Wireless Optical Channel for Underwater Visible Light Communication

The  growing  requirements for  underwater  observation  and  subsea monitoring systems have stimulated considerable interest in advancing the enabling technologies of underwater wireless communication and underwater sensor  networks.  This communication technology is expected to play an important role in investigating climate change, in monitoring biological, biogeochemical, evolutionary, and ecological changes in the sea, ocean, and lake environments. It would also help to control and maintain oil production facilities and harbors using unmanned underwater vehicles (UUVs), submarines, ships, buoys, and divers. However, the current and existing technology of underwater acoustic communication cannot provide the high data rate required to investigate and monitor these environments and facilities. Optical wireless communication especially visible light communication (VLC) has been proposed as the best alternative to meet this challenge.  In this work, characterization of wireless optical channel for underwater VLC will be carried out; models for VLC links will be developed to ensure high data rate for the reasonable distance of few tens of meters. Interesting result is expected which can be of very high importance for such applications.

Electromagnetic Simulation of Foliage and Camouflage Penetrative Imaging

During military operations against the enemy in the forest areas and also during the counter insurgency operations, identification of the objects (human beings, animals and weapon systems)) behind thick foliage or camouflage becomes absolutely necessary to ensure the success of the operations. However, Foliage Penetrative Imaging is still a developing field and there are only a small number of open publications which can be adapted to cost effective real world applications [1,2].

In the proposed project, fundamental studies on the following aspects will be undertaken using the simulation software like Ansys HFSS :

  1. absorption / transmission / reflection characteristics of electromagnetic waves of different frequencies through different types and densities of  foliage.
  2. absorption / transmission / reflection of electromagnetic waves of different frequencies on human and animal bodies and other materials, which are hidden behind the foliage / camouflage.
     

The aim is to identify those frequency bands, which are more effective in penetrating the foliage / camouflage and in detecting and identifying the objects (human beings, animals and military weapon systems) behind thick foliage or camouflage.

References

[1]  F. Amato, A. Farina, M. Fiorini & S. Gallone, “Surveillance Unattended Foliage Penetrating Radar for Border Control and Homeland Protection”, TransNav, the International Journal on Marine Navigation and Safety of Sea Transportation, Volume 7, Number 2, pp 193-196, June 2013
[2]  Lei Pengzheng, Huang Xiaotao, Fan Chongyi, Ji Kefeng, Qin Xianxiang, “Moving human target detection in foliage environments based on Hough transform”, IEEE IGRASS 2013, pp 2637-2640, 2013

Electromagnetic Simulation of Real-Time Through-the-Wall Imaging

Identification of the objects (including human beings and animals) behind an opaque wall using Through Wall Imaging (TWI) is a promising technology for many applications like security, military and search and rescue missions [1]. However, Through-Wall Imaging (TWI) is still a developing field and there are few open publications which can be adapted to cost effective real world applications. For instance, many of them uses extremely time consuming computations [2-5] or expensive hardware [6,7] which restricts their usage in real time applications, such as search and rescue. 

In the proposed project, fundamental studies on the following aspects will be undertaken using the simulation software like Ansys HFSS :

  1. absorption / transmission / reflection characteristics of electromagnetic waves of different frequencies through different wall material and thickness.
  2. absorption / transmission / reflection of electromagnetic waves of different frequencies on human and animal bodies and other materials.
     

The aim is to identify those frequency bands, which are more effective in detecting and identifying the objects (mainly human beings and animals) behind an opaque wall using Through Wall Imaging (TWI).

References

[1] E. J. Baranoski, “Through-wall imaging: Historical perspective and future directions,” Journal of the Franklin Institute, vol. 345, pp. 556–569, January 2008.
[2] C. Debes, J. Hahn, A. Zoubir, and M. Amin, “Target discrimination and classification in through-the-wall radar imaging,” IEEE Transactions on Signal Processing, vol. 59, pp. 4664–4676, Oct 2011.
[3] G. Smith and B. Mobasseri, “Robust through-the-wall radar image classification using a target-model alignment procedure,” IEEE Transactions on Image Processing, vol. 21, pp. 754 – 767,  Feb 2012.
[4] Y.-S. Yoon and M. Amin, “Compressed sensing technique for high-resolution radar imaging,” in Proceedings of SPIE, vol. 6968, p. 69681A, 2008.
[5] M. Leigsnering, C. Debes, and A. M. Zoubir, “Compressive sensing in through-the-wall radar imaging,” in IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP), 2011.
[6] T.S. Ralston, G.L. Charvat, and J.E. Peabody, "Real-time Through-wall Imaging Using an Ultra- Wideband (UWB) Multiple-Input Multiple Output (MIMO) Phased Array Radar System,"Proceedings of IEEE International Symposium on Phased Array Systems and Technology,October 2010, pp. 551–558.
[7]Yunqiang Yang, Ceming Zhang, and Aly E. Fathy, " Development and Implementation of Ultra-Wideband See-Through-Wall Imaging System Based on Sampling Oscilloscope,"IEEE  Antennas and Wireless Propagation Letters, vol. 7, 2008, pp.465-468.

Electronic System Level Design - study and implementation using open source approach

With the traditional boundaries between software and hardware disappearing and merging into system level, the hardware and software design flows are moving up the chain to Electronic System Level Design (ESLD). ESLD flow and methodology is still evolving and even the tools and technologies required to enable ESLD are getting established. It will be very interesting and challenging to study the current scenario of ESLD flows, methodologies, languages, tools and technologies and come up with a new flow and implement the same using the open source approach. This project envisages a thorough literature survey, designing ESLD flow and implementing the same on a large application like H.264 decoder, Face Detection/Recognition system, Channel Decoder in wireless communication, using SystemC and/or System Verilog/e language using open source tools.

Emotion Recognition from Facial Expressions

Objective:

  1. To design a robust multimodal (face and speech) feature set to characterize six basic emotions efficiently under constraints like variation in pose and illumination.
  2. To identify a suitable classifier to classify these emotions accurately
     

Project Summary:

Facial expressions, speech variations and recognizing emotions are an important factor in human-human interaction. It leads researchers on embodied conversational agents and social robots to exploration of polite behavior design to engage with humans. Automatic emotion recognition deals with identifying emotional state of the speaker from image/voice signal.

The proposed project features cognitive system and an emotion system. Both operate in parallel and are deeply intertwined to foster appropriately adaptive functioning of the robot in the environment as it interacts with people.

The current work would involve emotion recognition under pose & illumination variations

Vision based Navigation and Control of Biped Robot on Uneven Surfaces

Walking of a Biped Robot on normal surface has been investigated since two decades. Walking on uneven terrain poses difficult decision and stability concerns for a biped robot. Incorporating sensors for mapping the terrain is still in the early stages of development. Developing a stable walking strategy for a biped robot by combining the terrain information and control takes humanoid robotic research to a new level. Also it has immense practical applications in real world.

The research findings can be used in real world applications in the field of defence, service industries and in hazardous environments for replacing the human workforce.
The objective of the project is to develop, design and implement a real-time feedback control system for a biped robot for a stable and smooth movement over any given terrain including uneven surfaces. This project aims to :-

  1. Generate a stable reference trajectory which the biped can follow
  2. Develop a feedback controller for tracking the above reference trajectory and
  3. Finally to implement the controller and sensor on a biped robot enabling it to move smoothly over any uneven surface in real time.
     

The preliminary work involves the development of the controller for tracking a stable reference trajectory generated off-line using software simulation studies. The trajectory will be generated after considering the stability constraints using Zero Moment Point analysis. Once the controller is successfully designed, the mapping of terrain information using vision sensors and online calculation of the stable trajectory will be taken up. The machine vision module and the controller module will be integrated together to get a complete model which can help a biped robot to navigate smoothly over any uneven terrain.

207
PROGRAMS
OFFERED
6
AMRITA
CAMPUSES
15
CONSTITUENT
SCHOOLS
A
GRADE BY
NAAC, MHRD
8th
RANK(INDIA):
NIRF 2018
150+
INTERNATIONAL
PARTNERS