A5023610823- Advanced Neural Network Applications
- Sinusitis and nasal conditions
- Brain Tumor Detection and Classification
- Advanced Image Fusion Techniques
- AI in cancer detection
- Synthesis and Characterization of Heterocyclic Compounds
- Smart Systems and Machine Learning
- Synthesis and biological activity
- Anomaly Detection Techniques and Applications
- Image and Signal Denoising Methods
- Advanced Image Processing Techniques
- EEG and Brain-Computer Interfaces
- Antifungal resistance and susceptibility
- Adversarial Robustness in Machine Learning
- Remote Sensing and Land Use
- Currency Recognition and Detection
- Infectious Diseases and Mycology
- Image and Video Quality Assessment
- Head and Neck Surgical Oncology
- ECG Monitoring and Analysis
- Energy, Environment, and Transportation Policies
- Artificial Intelligence in Healthcare
- Image Enhancement Techniques
- Machine Learning and Data Classification
- Sports Dynamics and Biomechanics
Chitkara University
2023-2025
Dr. B. R. Ambedkar National Institute of Technology Jalandhar
2024
Institute of Engineering
2023
IFTM University
2023
Delhi Technological University
2021-2022
Banaras Hindu University
2022
Institute of Medical Sciences
2022
Indian Institute of Technology Hyderabad
2020
DAV University
2016
The capability of the self-attention mechanism to model long-range dependencies has catapulted its deployment in vision models. Unlike convolution operators, offers infinite receptive field and enables compute-efficient modeling global dependencies. However, existing state-of-the-art attention mechanisms incur high compute and/or parameter overheads, hence unfit for compact convolutional neural networks (CNNs). In this work, we propose a simple yet effective "Ultra-Lightweight Subspace...
The number of groups ($g$) in group convolution (GConv) is selected to boost the predictive performance deep neural networks (DNNs) a compute and parameter efficient manner. However, we show that naive selection $g$ GConv creates an imbalance between computational complexity degree data reuse, which leads suboptimal energy efficiency DNNs. We devise optimum size model, enables balance cost movement cost, thus, optimize energy-efficiency Based on insights from this propose "energy-efficient...
Reducing the influence of significant noise signal components on obtained raw ECG is essential for precise identification cardiac arrhythmias (CA), which frequently present as irregularities in heart rate or rhythm. Preprocessing used to remove signals and baseline drift from wave that recorded using internet things (IoT). After that, denoised subjected dimensionality reduction feature extraction. In order determine whether classification method more effective detecting arrhythmias, this...
Mucormycosis is an opportunistic infection which caused by fungus of the order Mucorales most common one Rhizopus oryzae. These infections are more in patients suffering from diabetes mellitus, malignancy, burn, severe trauma, malnutrition, renal failure, prolong neutropenia, immunosuppressed, long term steroid therapy or immunosuppressive therapy, hematopoietic stem cell transplant and solid organ recipients. Patients with serious illness 10 times prone to develop bacterial fungal secondary...
Fungal rhinosinusitis (FRS) once considered a rare disease. This global rise in the burden of fungal disease is consequence an increment population with weakened immune systems. Increased life expectancy conditions like diabetes mellitus, medical advancements invasive interventions, use steroid, wider uses broad-spectrum antibiotics, immunosuppressive treatments for transplantation and autoimmune diseases, increased incidence deficiency infections paranasal sinuses are fact spectrum diseases...
Knowledge distillation (KD) is generally considered as a technique for performing model compression and learned-label smoothing. However, in this paper, we study investigate the KD approach from new perspective: its efficacy training deeper network without any residual connections. We find that most of cases, non-residual student networks perform equally or better than their versions trained on raw data (baseline network). Surprisingly, some they surpass accuracy baseline even with inferior...
As they represent a sizable class of naturally occurring and synthetic chemicals with potent biological activity in the pharmaceutical fields, 1, 3 Thiazine heterocycles are great interest. The goal this project was to create Schiff base. 3,-Methoxychalcone derivative 1,3-thiazine. TLC, IR, 1HNMR, 13C NMR, each mass calculation were accustomed identify compositions recently created targeted substances. Using an Elevated plus maze, test compounds (B1–11) examined for their ability reduce...
Image recovery algorithms are used to enhance the advent of virtual pix. Those help picture by casting off low-frequency noise, polishing edges, and minimizing degradation because camera motion blur, movement aberration, aerial distortion. recuperation utilized in processing programs, including scientific imaging, satellite for pc object reputation, competent evaluation. This paper explores software healing photograph improve accuracy exceptional end product. Unique gain desired results....
This study examines the effect of various image recovery algorithms on photo great in digital processing. Especially, consequences visual quality photographs are studied through diverse metrics along with contrast, brightness, sharpness, and color saturation. Additionally, compared to every other as a way decide which algorithm produces effects. The studies will use images from public picture databases effects can be gathered first-rate evaluation software. Moreover, noise distortion...
The present-day state of real-time statistics analysis, in the main, relies on guide evaluation techniques that require trained records scientists as a way to extract significant insights from streams. Leveraging machine-gaining knowledge (ML) methods provides capacity for increasing efficiency and accuracy analysis. This paper presents an automation process data utilizing machine learning algorithms improve proposed involves preprocessing, feature selection, algorithm model training,...
In recent years, there has been a lot of interest in categorizing sleep stages using continuous photoplethysmography, or PPG data. This study presents powerful deep learning strategy for classifying various stages. The algorithm was trained proposed solution outperforms current approaches detecting phases because it employs three-stage approach that incorporates data preprocessing, feature extraction, and the development model. model's accuracy reliability were further evaluated by expertly...