Thesis Supervision

Dr. Kangkana Bora

Institute Post Doctoral Fellow.  Research area: Biomedical Image Processing.  

Research objectives: One important  objective is  to propose an intelligent system for automatic categorization of pap smear images to detect cervical dysplasia. An endoscope is a medical instrument that acquires images inside the human body. Another objective is to propose a new approach for  automatic detection of polyp regions in an endoscope image using  machine learning approaches. (February, 2019 to August, 2019)

Email: kangkana.bora@iitg.ac.in

Syuw Kang

This research work focuses on applying intelligent Closed-Circuit Television (CCTV) systems to strengthen counter-terrorism capabilities and improve the security of mass transportation networks. Conducted jointly with the University of Queensland (UQ), the project aims to develop advanced video analytics that enable early threat detection and rapid response. (Jointly in UQ)  (Defended on 02.05. 2007)

Thesis Title: Tracking of Persons for Intelligent CCTV-based Video Surveillance.

Malathi. T

Stereo correspondence finds corresponding matching pixels in the stereo image pairs. The difference between the coordinates of these matching pixels gives the disparity value, which in turn can be used for finding the depth information of a scene. The thesis deals with the problems of finding accurate disparity map. (Defended on 24. 07. 2017)

Thesis Title: Estimation of Disparity Map from Stereo Image Pairs in Presence of Occlusion

Email: malathi@iitg.ac.in

Sunil Kumar

Recognition of human’s emotion through facial expressions has many important applications. This research work aims at extracting facial informative regions and  discriminative shared space for facial expression recognition. (Defended on 06. 11. 2017).

Thesis Title: Extraction of Facial Informative Regions and Discriminative Shared Space for Facial Expression Recognition

Email: snk@iiitm.ac.in

Biplab K. Chakraborty

Skin detection is an important step in various image processing and vision-based Human  Computer  Interaction  (HCI) applications. Skin detection is the process of finding skin-coloured pixels and regions in an image or a video. A set of skin detection algorithms is proposed for different imaging conditions using chromatic and textural properties of skin regions. (Defended on 15. 03. 2018).

Thesis Title: A Novel Framework for Segmentation of Skin Regions using Chromatic and Textural Information

Email: ketanbiplab@gmail.com

Amit Vishwakarma

Image fusion is a technique to assimilate information acquired from similar or dissimilar sources of images. Compared to the source images, fused images convey more information, and they have more clarity. Different image fusion methods are proposed in the thesis. (Defended on 27. 07. 2019).

Thesis Title: Multi-sensor Image Fusion using Optimized and Adjustable Non-subsampled Shearlet Transform and Measurement of Fusion Performance.

Email: a.vishwakarma@iitg.ac.in

Tilendra Choudhary

Cardiac diseases are a leading cause of death globally, making continuous monitoring essential. Seismocardiogram (SCG) signals, captured non-invasively from chest vibrations, provide valuable insights into cardiac mechanics. This study focuses on estimating SCG fiducial points for accurate diagnosis, aiding personalized healthcare, telemonitoring, and body area networks. Various standalone and cardiac-assisted methods for SCG delineation are explored. (Jointly with Dr. L.N. Sharma.) (Defended on 19. 10. 2020). 

Thesis Title: Waveform Delineation and Analysis of Seismocardiographic Signals

Email: tilendra@iitg.ac.in

Pradipta Sasmal

Early and accurate diagnosis of colorectal cancer (CRC) is vital for effective treatment. Since polyps are precursors to CRC, colonoscopy is used to detect, segment, and classify them. However, manual analysis of colonoscopy video frames is time-consuming and prone to error due to visual similarities and subtle features. This thesis proposes automated frameworks for polyp detection, segmentation, and classification, enabling virtual biopsy for early dysplasia diagnosis. The methods aim to support clinicians by improving accuracy, efficiency, and enabling telemonitoring in medical settings. (Defended on 20. 08. 2022).

Thesis Title:  Automated Detection and Classification of Polyps in Colonoscopy Videos

Email: s.pradipta@iitg.ac.in

Debajit Sarma

In this research, we focus on the recognition of dynamic hand gestures that are primarily characterized by global motion along with distinct spatio-temporal and motion features. These are referred to as trajectory-based gestures. This dissertation investigates various methods for recognizing trajectory-based gestures and proposes novel models to enhance their recognition accuracy and efficiency. (Defended on 21. 10. 2022).  

Thesis Title: Hand Detection and Segmentation Schemes for Trajectory-guided Gesture Recognition

Email: s.debajit@iitg.ac.in

Anjan Kumar Talukdar

Gesture segmentation is a prerequisite stage to continuous gesture recognition which locates the start and end points of a gesture in an input sequence.  In this work, we focus our attention on coping with this problem associated with continuous gesture recognition. (Defended on 27.06.2022)

Thesis Title: Movement Epenthesis Detection in Compressed and Uncompressed Continuous Sign Language Videos.

Email: anjantalukdar@gauhati.ac.in

Pranab Jyoti Haloi

Content-based visual information retrieval (CBVIR) or content-based image retrieval (CBIR) has been one on the most vivid research areas in the field of computer vision. This research works aims at developing novel algorithms for CBIR. (Defended on 15.05. 2024) 

Thesis Title: Story Segmentation and Retrieval of News Videos in a Multi-modal Framework

Email: pranabjyoti2003@gmail.com

Nadeem Atif

The thesis develops a suite of six ultra-lightweight semantic segmentation networks that systematically improve the accuracy–efficiency trade-off for autonomous driving on edge devices. It introduces novel architectural modules—such as CAP, HAND-SPP, TAP, SLICE, DEDUCE, CTP, and DECoDe—that enhance multi-scale context modeling, preserve fine details, and enable real-time or near real-time performance on embedded GPUs and FPGA platforms. Together, these models demonstrate that high-precision semantic segmentation can be achieved with extremely compact, hardware-friendly designs suitable for practical deployment in resource-constrained environments. (Date thesis submission 18.11.2025) (Jointly with Prof. S.R. Ahmed).

Thesis Title: Design and Development of Semantic Segmentation Networks for Autonomous Driving on Embedded Platforms.

Email: atif176102103@iitg.ac.in

H. Pallab Jyoti Dutta

Hand gestures are a natural and widely used form of human communication, essential for seamless human-computer interaction. However, accurate gesture recognition is challenged by factors like background clutter, lighting variations, occlusion, and diverse hand appearances. While existing research has addressed these issues, generalization remains a challenge. This dissertation aims to develop a robust hand gesture recognition method and apply it to create user-centered interaction interfaces. (Defended on 18.12.2024)

Thesis Title: Hand Gesture Detection and Recognition for Gesture-based Patient Rehabilitation and Assistance Systems.

Email: h18@iitg.ac.in

Mousumi Das

Many vital parameters, such as heart rate (HR), heart rate variability (HRV), blood pressure (BP), oxygen saturation (SpO2), and respiratory rate (BR) provide insight to cardiac health and help in diagnosing and treating life-threatening diseases. In this study, two emerging cardiac modalities, such as seismocardiography (SCG) and remote photoplethysmography (rPPG) are considered for the estimation of cardiac vital parameters.  (Jointly with Dr. L.N. Sharma) (Defended on 04.02.2025)

Thesis Title: Cardiac Parameters Estimation Using Seismocardiographic and Remote Photoplethysmographic Signals

Email: mousumi18a@iitg.ac.in

Allen Patnaik

Single-image super-resolution (SISR) has emerged as a critical research area due to its significance in enhancing the quality of both natural and remote sensing images for a wide range of applications. This thesis focuses on reusing key features and effectively capturing fine-grained details along with the global context of images to achieve high-quality image reconstruction.  (Defended on 31.10.2025)

Thesis Title: Single image super-resolution for remote sensing images – adversarial and multibranch deep architectures with different attention mechanisms

Email: allen.patnaik@iitg.ac.in

Vanshali Sharma

This thesis tackles challenges in analyzing colorectal cancer (CRC) screening videos, including low-quality data and limited annotations. It proposes an automated pipeline that mimics clinical workflows to enhance diagnostic accuracy and reduce human error. (Jointly with Prof. P.K. Das, Dept. of CSE).  (Defended on 27.08.2024)

Thesis Title: Enhancing Anomaly Detection in Endoscopy Images and Videos: From Keyframe Extraction to Clinical and Synthetic Dataset Design

Email: vanshalisharma@iitg.ac.in

Saquib Mazhar

This thesis enhances real-time semantic segmentation for autonomous driving by introducing a composite loss to improve small object accuracy and proposing three lightweight, GPU-efficient models—IRDPNet, BANet, and CGMANet—that reduce complexity while maintaining high segmentation performance through efficient feature extraction and context-aware attention. (Defended on 29.05.2025) (Jointly with Prof. S.R. Ahmed).

Thesis Title: Lightweight Deep Learning Architectures for Semantic Scene Segmentation for Applications in Autonomous Driving.

Email: saquibmazhar@iitg.ac.in

Chaudhari Harshal Rajendra

This work advances off-axis DHM for label-free, high-res bio-imaging by using FRFT for better reconstruction, SVD for aberration correction, and a CNN-ViT model for cell classification. A compact, shearing-based DHM setup simplifies optics, while deep learning enables fast, high-throughput holographic flow cytometry with improved autofocusing. (Defended on 27.11. 2025) (Jointly with Dr. Rishikesh Dilip Kulkarni).

Thesis Title: Digital Off-Axis Holographic Microscopy for Cell Imaging and Classification

Email: charshal@iitg.ac.in

Bibek Goswami

Early Diagnosis of Oral Cancer Using Multiple Imaging Modalities (Date of joining 27.07.2020)

Thesis Title: Medical Image Analysis.

Email: b.goswami@iitg.ac.in

Riyaj Uddin Khan

Biomedical signal processing, machine learning, brain-computer interface, neurodegenerative diseases, computer-aided medical diagnosis, and Alzheimer's disease (AD). (Date of joining 19.07. 2019 & continuing.) (Jointly with Prof. S.R. Ahmed).

Thesis Tiltle: Brain-computer interface and neurodegenerative diseases. 

Email: r.uddin@iitg.ac.in

Kamal

PhD Research Scholar (Mehta Family School of Data Science and Artificial Intelligence),  Computer Vision (Date of joining 28.07. 2021 & continuing.) (Jointly with Dr. D. R. Neog)

Thesis Title: Computer Vision. 

Email: kkamal@iitg.ac.in

Nitin Yadav

Computer Vision (Date of joining 04.08. 2022 & continuing.)

Thesis Title: Computer Vision. 

Email: nitin.yadav@iitg.ac.in

Pallapu Mohan Krishna

PhD Research Scholar (Mehta Family School of Data Science and Artificial Intelligence), Computer Vision (Date of joining 26.07. 2022 & continuing.) (Jointly with Dr. D. R. Neog)

Thesis Title: Computer Vision. 

Email: k.pallapu@iitg.ac.in

Dipima Sarma

PhD Research Scholar (Employed Part-time), Centre for Disaster Management and Research (CDMR), Jointly with Prof. Arup Kumar Sarma, Department of Civil Engineering, IIT Guwahati. (Date of Joining: 26.07.2022)  

Thesis Title: Remote Sensing

Email: s.dipima@iitg.ac.in

Gautam Talukdar

Image processing (Date of joining 27.12. 2023 & continuing.) (QIP)

Thesis Title: Image Processing

Email: gautam.talukdar123@gmail.com

Sanjeev Kumar

HCI & Computer Vision (Date of joining 02.01. 2024 & continuing)

Thesis Title: HCI, Computer Vision. 

Email: sanjeevk2117@iitg.ac.in

Neha Spriha Baruah

Computer Vision (Date of joining 02.08. 2024 & continuing.) Thesis Title: Computer Vision.     

Email: neha.baruah@iitg.ac.in

Ashish Kumar

Computer Vision (Date of joining 02.08. 2024 & continuing.) Thesis Title: Computer Vision.     

Email: ashishk2002@iitg.ac.in

Himasree Deka

VLSI and Machine Learning (Date of joining 02.08. 2024 & continuing.) (Jointly with Prof. S.R. Ahmed).

Email: d.himasree@iitg.ac.in

Deep Haloi

Computer Vision (Date of joining 21.07. 2025 & continuing.) Thesis Title: Computer Vision.    

Email: d.haloi@iitg.ac.in

Mridupaban Bordoloi

Machine Learning & HCI (Date of joining 21.07. 2025 & continuing.) (Jointly with Dr Samit Bhattacharya, CSE). 

Email: b.mridupaban@iitg.ac.in