A5078638232- Speech and Audio Processing
- Wireless Body Area Networks
- Antenna Design and Analysis
- Emotion and Mood Recognition
- Music and Audio Processing
- Cardiac Imaging and Diagnostics
- Energy Harvesting in Wireless Networks
- Corneal Surgery and Treatments
- Photonic and Optical Devices
- GaN-based semiconductor devices and materials
- Semiconductor materials and interfaces
- Cardiovascular Function and Risk Factors
- Hematological disorders and diagnostics
- Stock Market Forecasting Methods
- Hand Gesture Recognition Systems
- Wireless Power Transfer Systems
- Quantum Dots Synthesis And Properties
- Cardiac, Anesthesia and Surgical Outcomes
- Linguistic Studies and Language Acquisition
- Blockchain Technology Applications and Security
- Speech Recognition and Synthesis
- Advanced X-ray and CT Imaging
- Semiconductor Quantum Structures and Devices
- Nanowire Synthesis and Applications
- Ocular Oncology and Treatments
Massachusetts General Hospital
2024-2025
American InterContinental University
2024
American University
2024
Noakhali Science and Technology University
2024
Tel Aviv Sourasky Medical Center
2024
Tel Aviv University
2024
University of Bedfordshire
2024
Shahjalal University of Science and Technology
2010-2023
Ahsanullah University of Science and Technology
2023
National Institute of Nuclear Medicine & Allied Sciences
2018-2023
In this study, we have presented a deep learning-based implementation for speech emotion recognition (SER). The system combines convolutional neural network (DCNN) and bidirectional long-short term memory (BLSTM) with time-distributed flatten (TDF) layer. proposed model has been applied the recently built audio-only Bangla emotional corpus SUBESCO. A series of experiments were carried out to analyze all models discussed in paper baseline, cross-lingual, multilingual training-testing setups....
In this paper, we propose a new method for Bengali Sign Language Recognition using Deep Convolutional Neural Networks (DCNN). The is built to recognize static hand signs of 37 letters the alphabet. Conducting tests on three sets (we have used 31 images different signs) with total 1147 images, varying precision feature points taken each test. With use deep convolutional neural network and utilizing learned features from pre-trained fine-tuning top layers network, achieved high overall...
SUBESCO is an audio-only emotional speech corpus for Bangla language. The total duration of the in excess 7 hours containing 7000 utterances, and it largest available this Twenty native speakers participated gender-balanced set, each recording 10 sentences simulating seven targeted emotions. Fifty university students evaluation corpus. Each audio clip corpus, except those Disgust emotion, was validated four times by male female raters. Raw hit rates unbiased were calculated producing scores...
Background: The association between type 2 diabetes mellitus (T2DM) and bone mineral density (BMD) remains unclear. Screening, identification prevention of potential risk factors for osteoporosis in T2DM patients are important necessary preserving a good quality life decreasing the fracture. This study is expected to help us find trend BMD Bangladesh identify any that may predispose low BMD. Objective: To evaluate with long term history type-2 mellitus. Procedure: cross sectional...
Centralized marketplaces run on a process where single authority has control over all activity. They suffer from limitation such as- mandatory fees to be paid the platform for listing and selling items, lack of privacy account, improper transaction security so on. This paper presents decentralized application aiming reduce these drawbacks using Ethereum blockchain. It was implemented truffle evolvement framework with solidity language. smart contracts contain functions this application....
This study presents a comprehensive analysis of machine learning techniques for measuring and predicting security in stock markets, comparing the performance supervised, unsupervised, deep models. Using diverse dataset from Kaggle that includes historical prices, financial news sentiment, company fundamentals, macroeconomic indicators, we applied feature engineering rigorous preprocessing methods to optimize model accuracy. The evaluated Random Forest, Support Vector Machines (SVM), K-Means...
In this paper, we propose a method for Bengali Sign Language Recognition based on skin segmentation and geometric hashing. The area is obtained using combination of dynamic color-based thresholding mean shift the original image. A novel feature extraction algorithm introduced which tries to identify hand by placing points at regular intervals along perimeter blob. dataset 1147 images also prepared task training hash table map with co-ordinates points. built recognize static signs 37...
In this paper, a highly sensitive miniaturized surface plasmon resonance (SPR) based photonic crystal fiber (PCF) sensor is presented for wide range of analyte sensing. Gold selected as the plasmonic metal its higher chemical stability and titanium oxide works adhesive layer gold attachment on silica. The sensing medium are placed exterior to design make it fitting practical applications. By careful arrangement periodic refractive index in design, generation evanescent fields fine-tuned...
This paper presents a brief study of remarkable works done for the development Automatic Speech Recognition (ASR) system Bangla language. It discusses information available speech corpora this language and reports major contributions made in research paradigm last decade. Some important design issues to develop recognizer are: levels recognition, vocabulary size, speaker dependency approaches classifications; these have been defined order complexity recognition. also highlights on some...
In this age of miniaturization and nanotechnology, nano sized zero dimensional particles called quantum dots are finding wide range applications. The reasons behind their growing popularity controllable electrical optical properties. This control is achieved by tuning bandgaps changing shape size. But effects most prominent only within strong confinement regime. That is, the radius much less than exciton Bohr bulk material which made. paper, we have studied on Indium Gallium Nitride,...
In this paper, A standardized and implantable planar inverted f antenna is accomplished with modified design, featuring balanced size miniaturization for wireless communications. This has an intensified reliability feasibility in turns performing biomedical application. The proposed represented For MICS BAND. band contains 402 to 405 MHz resonant frequency. been chosen because of its tractability towards higher bandwidth the decreasing antenna. 12.5 × 12.5mm2 miniature dimension offered...
A modified design and implantable planar inverted F-antenna is examined in this paper. This structure can be applied to improve antennas for various implantation scenarios biomedical applications. The proposed antenna designed human head (scalp) phantom analyzed the performance bio-environment. flexible PIFA operated at Medical Implant Communication Service (MICS) Band (402–405 MHz). MICS band has been chosen because of its tractability towards higher bandwidth with reducing size antenna. By...
This paper represents a standard design of Planar Inverted-F Antenna (PIFA) for medical applications which operates in MICS band (402–405MHz). The antenna is designed very small size and better resonant performance observed at different tissues human phantom Model. Rogers R03010 used as substrate Copper material. By employing meandering including the superstrate, dimension proposed to 12.5 × mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Although physicians think that tuberculosis (TB) is a disease of the 19th century or limited to ethnic minorities and those with coexistent such as AIDS, TB remains us can strike any individual. The diagnosis, especially in early stages disease, may be problematic, if it not considered likely. However, older techniques Ga-67 citrate well have edge over newer when comes diagnosing this "old time" disease.
To better understand human behavior, it is essential to investigate the speech features that contribute most emotional expressions. In this study, we investigated how different emotions affect acoustic properties of speech. This study explored a new set widely utilized recognize from audios. Experimental investigation using Bangla and English datasets were conducted SVM, Random forest, XGBoost algorithm. We used Grid Search method with five-fold cross-validation select optimal parameters for...
Abstract Pneumonia is a bacterial or viral respiratory ailment that affects lot of people, especially in impoverished and newly industrialized countries due to pollution, unclean living conditions, overcrowding. The objective this research focuses on developing an automated approach for correctly identifying pneumonia from X-rays the chest images order decrease time needed manual evaluation. In study, two imbalanced datasets—the Corona Hack Chest X-Ray dataset X-ray Images dataset—were...