A5016222651- Face recognition and analysis
- Context-Aware Activity Recognition Systems
- Non-Invasive Vital Sign Monitoring
- Face and Expression Recognition
- Emotion and Mood Recognition
- Parallel Computing and Optimization Techniques
- Real-Time Systems Scheduling
- Video Surveillance and Tracking Methods
- EEG and Brain-Computer Interfaces
- Biometric Identification and Security
- ECG Monitoring and Analysis
- Speech and Audio Processing
- Human Pose and Action Recognition
- Distributed systems and fault tolerance
- Music and Audio Processing
- Heart Rate Variability and Autonomic Control
- User Authentication and Security Systems
- Hand Gesture Recognition Systems
- Video Analysis and Summarization
- Anomaly Detection Techniques and Applications
- Image and Video Quality Assessment
- Electron and X-Ray Spectroscopy Techniques
- Intelligent Tutoring Systems and Adaptive Learning
- Technology Use by Older Adults
- Pediatric Pain Management Techniques
Mitford Hospital, Dhaka
2024
Bangladesh University of Engineering and Technology
2008-2024
Dhaka Medical College and Hospital
2024
Aarhus University
2022-2023
Aalborg University
2013-2022
Mugda Medical College
2022
Comilla Medical College
2015-2022
George Mason University
2011-2016
Media Design School
2016
Jahangirnagar University
2015
On emerging multicore systems, task replication is a powerful way to achieve high reliability targets. In this paper, we consider the problem of achieving given target for set periodic real-time tasks running on system with minimum energy consumption. Our framework explicitly takes into account coverage factor fault detection techniques and negative impact Dynamic Voltage Scaling (DVS) rate transient faults leading soft errors. We characterize subtle interplay between processing frequency,...
Pain is a symptom of many disorders associated with actual or potential tissue damage in human body. Managing pain not only duty but also highly cost prone. The most primitive state management the assessment pain. Traditionally it was accomplished by self-report visual inspection experts. However, automatic systems from facial videos are rapidly evolving due to need managing robust and effective way. Among different challenges video data two issues increasingly prevalent: first, exploiting...
Heartbeat rate (HR) reveals a person's health. This article presents an effective system for measuring HR from facial videos acquired in more realistic environment than current systems' testing environments. The proposed method uses feature point-tracking that combines good to track with supervised descent overcome the limitations of currently available video-based systems, including unrealistic restriction subject's movement and artificial lighting during data capture. A face quality...
Abstract Angle-resolved photoemission spectroscopy (ARPES) is a powerful experimental technique to determine the electronic structure of solids. Advances in light sources for ARPES experiments are currently leading vast increase data acquisition rates and quantity. On other hand, access time most advanced instruments remains strictly limited, calling fast, effective, on-the-fly analysis tools exploit this time. In response need, we introduce ARPESNet, versatile autoencoder network that...
In this paper, we present an energy-aware standby-sparing technique for periodic real-time applications. A system consists of a primary processor where the application tasks are executed using Dynamic Voltage Scaling (DVS) to save energy, and spare backup at maximum voltage/frequency, should there be need. our framework, employ Earliest-Deadline-First (EDF) Earliest-Deadline-Late (EDL) scheduling policies on CPUs, respectively. The use EDL CPU allows delaying as much possible, enabling...
Face recognition has attracted increasing attention due to its wide range of applications, but it is still challenging when facing large variations in the biometric data characteristics. Lenslet light field cameras have recently come into prominence capture rich spatio-angular information, thus offering new possibilities for advanced systems. This paper proposes a double-deep learning framework field-based face recognition, which able model both intra-view/spatial and inter-view/angular...
Physical fatigue reveals the health condition of a person at, for example, checkup, fitness assessment, or rehabilitation training. This study presents an efficient non‐contact system detecting non‐localised physical from maximal muscle activity using facial videos acquired in realistic environment with natural lighting where subjects were allowed to voluntarily move their head, change expression, and vary pose. The proposed method utilises feature point tracking by combining ‘good track’...
Energy and reliability management are important design constraints for real-time embedded systems. We consider the problem of achieving a given target set periodic applications running on multi-core system with minimum energy consumption. Specifically, we observe that emerging multicore platforms provide ample opportunities to use task replication achieve targets mitigate negative impact Dynamic Voltage Scaling (DVS) rate transient faults leading soft errors. However, while it allows using...
In this paper, we investigate the issues associated with facial expression recognition of Traumatic Brain Insured (TBI) patients in a realistic scenario.These have restricted or limited muscle movements reduced expressions along non-cooperative behavior, impaired reasoning and inappropriate responses.All these factors make automatic understanding their more complex.While existing systems showed high accuracy by taking data from healthy subjects, performance is yet to be proved for real TBI...
Facial dynamics can be considered as unique signatures for discrimination between people. These have started to become important topic since many devices the possibility of unlocking using face recognition or verification. In this work, we evaluate efficacy transition frames video in emotion compared peak identification. For experiments with extract features from each frame a fine-tuned VGG-Face Convolutional Neural Network (CNN) and geometric facial landmark points. To model temporal...
This research enhances crowd analysis by focusing on excessive and density predictions for Hajj Umrah pilgrimages. Crowd usually analyzes the number of objects within an image or a frame in videos is regularly solved estimating generated from object location annotations. However, it suffers low accuracy when far away surveillance camera. proposes approach to overcome problem taken camera at distance. The proposed employs fully convolutional neural network (FCNN)-based method monitor...
Traditional still camera-based facial image acquisition systems in surveillance applications produce low quality face images. This is mainly due to the distance between camera and subjects of interest. Furthermore, people such videos usually move around, change their head poses, expressions. Moreover, imaging conditions like illumination, occlusion, noise may change. These all aggregate most detected images terms measures resolution, pose, brightness, sharpness. To deal with these problems...
In this paper, we consider the feasibility problem of a set real-time jobs which may be subject to fault burst during execution. A represents time interval multiple incur faults; hence recoveries needed. We show that determining system, last at most Δ units, is an NP-Hard even when exact position known priori. However, in practical occur any arbitrary and unpredictable time. develop analysis by assuming recovery strategy where, addition job end detected, all preempted tasks are also...
Energy management and reliability are two important design objectives for real-time embedded systems. Recently, the standby-sparing scheme that uses a primary processor spare has been exploited to provide fault tolerance while keeping energy consumption under control through DVS DPMtechniques. In this paper, we consider technique fixed-priority periodic tasks. We propose dual-queue mechanism which execution of backup tasks maximally delayed, as well online algorithms manage consumption. Our...
Face alignment in video is a primitive step for facial image analysis. The accuracy of the greatly depends on quality face frames and low faces are proven to cause erroneous alignment. Thus, this paper proposes system aware by using Supervised Decent Method (SDM) along with motion based forward extrapolation method. proposed first extracts from frames. Then, it employs assessment technique measure quality. If high, uses SDM landmark detection. corrects landmarks that detected SDM. Depending...
Acute arboviral infections like dengue have a significant negative socioeconomic and health impact on many tropical subtropical areas of the world. About 3.9 billion Individuals are at risk contracting virus Asia bears brunt that load. Bangladesh, other south-east Asian countries faced massive outbreak in 2019. This cross sectional study was done three tertiary care centers Dhaka, Bangladesh during this worst dengue. The conducted from 1st July to 31st December, 2019 with an aim describe...
How can we efficiently and scalably cluster high-dimensional data? The k -means algorithm clusters data by iteratively reducing intra-cluster Euclidean distances until convergence. While it finds applications from recommendation engines to image segmentation, its application is hindered the need repeatedly compute among points centroids. In this paper, propose Marigold ( for data), a scalable clustering in high dimensions. prunes distance calculations means of (i) tight distance-bounding...