A5018647189- Advanced Image Processing Techniques
- Image Processing Techniques and Applications
- Advanced Vision and Imaging
- EEG and Brain-Computer Interfaces
- Digital Imaging for Blood Diseases
- AI in cancer detection
- Medical Image Segmentation Techniques
- Advanced Neuroimaging Techniques and Applications
- Advanced MRI Techniques and Applications
- Cell Image Analysis Techniques
- Generative Adversarial Networks and Image Synthesis
- Image and Signal Denoising Methods
- Blind Source Separation Techniques
- Functional Brain Connectivity Studies
- Neural dynamics and brain function
- Radiomics and Machine Learning in Medical Imaging
- Advanced Image and Video Retrieval Techniques
- Medical Imaging Techniques and Applications
- Human Pose and Action Recognition
- Optical measurement and interference techniques
- Digital Media Forensic Detection
- Anomaly Detection Techniques and Applications
- Image Retrieval and Classification Techniques
- Face recognition and analysis
- Sparse and Compressive Sensing Techniques
Indian Institute of Technology Mandi
2016-2025
Pandit Deendayal Petroleum University
2024
All India Institute of Medical Sciences
2023
Government of Himachal Pradesh
2023
National Institute of Technology Karnataka
2017
Sardar Vallabhbhai National Institute of Technology Surat
2017
Medical Technologies (Czechia)
2015
Indian Institute of Technology Madras
2006-2013
University of North Carolina at Chapel Hill
2013
Indian Institute of Technology Kanpur
1993
Breast cancer is one of the most common in women worldwide. It typically diagnosed via histopathological microscopy imaging, for which image analysis can aid physicians more effective diagnosis. Given a large variability tissue appearance, to better capture discriminative traits, images be acquired at different optical magnifications. In this paper, we propose an approach utilizes joint colour-texture features and classifier ensemble classifying breast histopathology images. While...
With the arrival of several face-swapping applications such as FaceApp, SnapChat, MixBooth, FaceBlender and many more, authenticity digital media content is hanging on a very loose thread. On social platforms, videos are widely circulated often at high compression factor. In this work, we analyze deep learning approaches in context deepfakes classification scenarios demonstrate that proposed approach based metric can be effective performing classification. Using less number frames per video...
Computerized approaches for automated classification of histopathology images can help in reducing the manual observational workload pathologists. In recent years, like other areas, deep networks have also attracted attention image analysis. However, existing paid little exploring multilayer features improving classification. We believe that considering multi-layered is important as different regions images, which are turn at magnifications may contain useful discriminative information...
Microscopic examination of biopsy tissue slides is perceived as the gold-standard methodology for confirmation presence cancer cells. Manual analysis an overwhelming inflow highly susceptible to misreading by pathologists. A computerized framework histopathology image conceived a diagnostic tool that greatly benefits pathologists, augmenting definitive diagnosis cancer. Convolutional Neural Network (CNN) turned out be most adaptable and effective technique in detection abnormal pathologic...
Denoising Diffusion Models (DDMs) are widely used for high-quality image generation and medical segmentation but often rely on Unet-based architectures, leading to high computational overhead, especially with high-resolution images. This work proposes three NCA-based improvements diffusion-based segmentation. First, Multi-MedSegDiffNCA uses a multilevel NCA framework refine rough noise estimates generated by lower level models. Second, CBAM-MedSegDiffNCA incorporates channel spatial...
White matter bundle segmentation is crucial for studying brain structural connectivity, neurosurgical planning, and neurological disorders. Matter Segmentation remains challenging due to similarity in streamlines, subject variability, symmetry 2 hemispheres, etc. To address these challenges, we propose TractoGPT, a GPT-based architecture trained on streamline, cluster, fusion data representations separately. TractoGPT fully-automatic method that generalizes across datasets retains shape...
Under stereo settings, the twin problems of image superresolution (SR) and high-resolution (HR) depth estimation are intertwined. The subpixel registration information required for is tightly coupled to 3D structure. effects parallax pixel averaging (inherent in downsampling process) preclude a priori motion superresolution. These factors also compound correspondence problem at low resolution (LR), which turn affects quality LR estimates. In this paper, we propose an integrated approach...
Cervical cancer is the second most common cause of death among women worldwide, but it can be treated if detected early. However, due to inter and intra observer variability in manual screening, automating process need hour. For classifying cervical cells as normal vs abnormal, segmentation nuclei well cytoplasm a prerequisite. But relatively more reliable equally efficient for classification that cytoplasm. Hence, this paper proposes new approach based on selective pre-processing then...
Diabetes mellitus is a widespread chronic metabolic disorder that requires regular blood glucose level surveillance. Current invasive techniques, such as finger-prick tests, often result in discomfort, leading to infrequent monitoring and potential health complications. The primary objective of this study was design novel, portable, non-invasive system for diabetes detection using breath samples, named DiabeticSense, an affordable digital device early detection, encourage immediate...
Depth map sensed by low-cost active sensor is often limited in resolution, whereas depth information achieved from structure motion or sparse scanning techniques may result a point cloud. Achieving high-resolution (HR) low resolution (LR) densely reconstructing non-uniformly sampled are fundamentally similar problems with different types of upsampling requirements. The first problem involves uniform grid, the second type requires an non-uniform grid. In this paper, we propose new approach to...
Diabetes is a metabolic disorder often diagnosed late and requires continuous monitoring of blood glucose. We introduce GlucoBreath, user-centric, cost-effective, portable pre-diagnostic solution to address this global challenge. GlucoBreath addresses the urgent need for an accessible non-intrusive diabetes detection device, offering affordability, mobility, comfortable non-invasive testing, especially among economically weaker sections society. comprises (i) multi-sensor Internet Things...