Dr. Gladys Ngetich

Thank you Prof. Hamid Jafarnejad Sani, PhD for visiting NJIT and presenting your exciting research work. If anyone is interested in safety and controls with applications to UAV systems, look at recent papers by Prof. Sani!

29

1 Comment

Our recently accepted paper in the IEEE Transactions on Intelligent Transportation Systems (T-ITS) proposes a novel failure-aware reactive UAV delivery service composition framework for rerouting drone deliveries in urban areas when service failures occur. #drones #dronedelivery #droneservices #research #uavs #uavservices #uavdelivery

28

3 Comments

Mehmet Savasci presented our paper titled "PADS: Power Budgeting with Diagonal Scaling for Performance-Aware Cloud Workloads" at the recently held IEEE Conference on Green and Sustainable Computing (IGSC) conference in Austin, TX. This paper presents a new hardware-agnostic diagonal scaling technique for power management and power capping in compute clusters. Read it here: https://lnkd.in/eyeKsrz8 Abel S. Ahmed Ali El Din David Irwin #energyefficiency #cloudcomputing

26

3 Comments

Another research topic: the field of medical robotics and intelligent systems. A list of research areas conducted by our ECE faculty is included at the end of this post. Assistant Professor Dr. Mostafa Rahbar's most recent paper, titled 'Enhanced U-Net with GridMask (EUGNet): A Novel Approach for Robotic Surgical Tool Segmentation,' can be found at the link below. Great work Dr. Rahbar! Mostafa Daneshgar Rahbar, Ph.D. is also currently working on another project with George P Pappas Ph.D. and Nabih Jaber, Ph.D. on an intraoperative visual intelligence system, specifically “Real-Time Surgical Instrument Segmentation for Enhanced Surgical Monitoring.” https://lnkd.in/eiSmfJDA This study proposed enhanced U-Net with GridMask (EUGNet) image augmentation techniques focused on pixel manipulation, emphasizing GridMask augmentation. This study introduces EUGNet, which incorporates GridMask augmentation to address U-Net’s limitations. EUGNet features a deep contextual encoder, residual connections, class-balancing loss, adaptive feature fusion, GridMask augmentation module, efficient implementation, and multi-modal fusion. These innovations enhance segmentation accuracy and robustness, making it well-suited for medical image analysis. The GridMask algorithm is detailed, demonstrating its distinct approach to pixel elimination, enhancing model adaptability to occlusions and local features. A comprehensive dataset of robotic surgical scenarios and instruments is used for evaluation, showcasing the framework’s robustness. Specifically, there are improvements of 1.6 percentage points in balanced accuracy for the foreground, 1.7 points in intersection over union (IoU), and 1.7 points in mean Dice similarity coefficient (DSC). These improvements are highly significant and have a substantial impact on inference speed. The inference speed, which is a critical factor in real-time applications, has seen a noteworthy reduction. It decreased from 0.163 milliseconds for the U-Net without GridMask to 0.097 milliseconds for the U-Net with GridMask. ECE faculty research areas include: . Intelligent Transportation Systems . Smart Grid Power Line Communications . Artificial Intelligence . Intelligent Systems . Medical Robotics . Wireless Communication, Vehicle Communication Network, Sensor Networks . Speech Enhancement Techniques . Internet of Things . Digital Signal Processing . Embedded System and Software . Non-Destructive Optical Detection, Ultrasound Generation and imaging, Photoacoustic Imaging and Sensing . Augmented reality #technology #robotics #smart #AI

44

2 Comments

Check out the poster presented by PhD student, Ishfaq Aziz, at the 2nd USACM Thematic Conference on Uncertainty Quantification for Machine Learning Integrated Physics Modeling (UQ-MLIP 2024). This poster was titled "PHYSICS-INFORMED MACHINE LEARNING (PIML) FOR ELECTROMAGNETIC WAVE PROPAGATION AND MATERIAL PROPERTY ESTIMATION" and discusses a solution that integrates machine learning with physics to analyze EM wave propagation for microwave radar sensing. #PIML #PINN #Microwave_sensing #radar_sensing #parameter_estimation #physics_informed #nondestructive_Evaluation

118

1 Comment

I am super thrilled to announce in the quiet of the night that our literature review on darknet traffic analysis has been accepted for publication in IEEE Access! A big shout out to the lead author, javeriah Saleem, and her principal supervisor, A/Prof Rafiqul Islam. You can access the paper for free at IEEE Access (https://lnkd.in/gWP-98g7). The dark web, nestled beneath the deep web within the internet protocol stack, remains hidden from conventional search engines like Google, Bing, and Yahoo. It is fascinating to learn that the vast majority of the web, about 95%, is part of this dark and generally inaccessible web - using specialized software such as the Onion Router (Tor) and the Invisible Internet Project (I2P), includes unused address spaces designed to operate independently of the global internet network. Recent findings indicate that 94% of requests originating from the Tor browsing network (which is a primary tool used to access parts of the darknet) pose cybersecurity threats. The advanced privacy tools required to access the darknet pose significant challenges for law enforcement agencies in detecting and prosecuting cybercriminal activities. Our paper contributes to the field by: * Providing a comprehensive review of the literature from 2017 to 2023 through a systematic literature review (SLR). * Examining the taxonomy of darknet traffic analysis in detail. * Presenting an in-depth analysis of detection methods for identifying darknet traffic attacks. * Highlighting the challenges and complexities in darknet traffic analysis. We aim to answer critical research questions, including: * The main themes in darknet traffic analysis. * The process for darknet traffic detection and classification, including the datasets used, the impact of feature selection, common machine learning algorithms, and metrics for measuring classification accuracy. * The taxonomy of darknet traffic analysis for the detection of malicious activities. * The challenges and future research directions in darknet traffic analysis. What are you thinking? Planning to read the freely available paper? Join us in exploring the depths of the dark side of the internet's hidden layers! #CharlesSturtResearch

52

9 Comments

Taking advantage of the weekend to post :-) ECE research addresses some of the most diverse, vibrant, and contemporary issues. #Technology #AI Lawrence Technological University Robotic Automation &Intelligent Equipment Robotics Lawrence Technological University

8

1 Comment

Thank Saint Louis University School of Science and Engineering for sharing our work! This dual recognition, being named a finalist for both #GordonBell (general track) and #GordonBell for #Climate Modelling, highlights the critical role of the #PaRSEC task-based runtime system, co-developed with University of Tennessee, Knoxville and a key focus of my research, in advancing complex scientific applications. 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐆𝐖𝐀𝐒: a finalist for #GordonBell, in partnership with KAUST (King Abdullah University of Science and Technology), Massachusetts Institute of Technology, and NVIDIA. By implementing mixed-precision computations and relying on PaRSEC, Exascale GWAS significantly enhances Kernel Ridge Regression (KRR), a critical technique for analyzing genetic data, achieving a performance improvement of five orders of magnitude over the state-of-the-art CPU-only REGENIE, the current leading software for similar tasks. More details: https://lnkd.in/gEUKwYd6. 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐂𝐥𝐢𝐦𝐚𝐭𝐞 𝐄𝐦𝐮𝐥𝐚𝐭𝐨𝐫: a finalist for #GordonBell for #Climate Modelling, in partnership with KAUST (King Abdullah University of Science and Technology), NSF NCAR - The National Center for Atmospheric Research, University of Notre Dame, and NVIDIA. This exascale climate emulator is to address the growing computational and storage demands of high-resolution Earth System Models. By leveraging advanced numerical methods with the PaRSEC dynamic runtime system, the resulting pipeline, stochastically models spatio-temporal variations in climate data, boosts computational efficiency, greatly enhances the accuracy of climate emulators, and provides critical insights for understanding climate change and extreme weather patterns. More details: https://lnkd.in/gqs_QgHr. Many thanks to our 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐆𝐖𝐀𝐒 team (Hatem Ltaief, Rabab Alomairy, Ph.D., Jie Ren, Lotfi SLIM, Thorsten Kurth, Benedikt Dorschner, Salim Bougouffa, Rached Abdelkhalek, and David Keyes), our 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐂𝐥𝐢𝐦𝐚𝐭𝐞 𝐄𝐦𝐮𝐥𝐚𝐭𝐨𝐫 team (Sameh Abdulah, Allison Baker, George Bosilca, Stefano Castruccio, Marc Genton, David Keyes, Zubair Khalid, Hatem Ltaief, Yan Song, Gera Stenchikov, Ying Sun), and the 𝐏𝐚𝐑𝐒𝐄𝐂 team (George Bosilca, Thomas Herault, Aurelien Bouteiller, and Joseph Schuchart)!

54

10 Comments

Recently published is our (Himan Shahabi · Mehdi Gholamnia · Jahanbakhsh Mohammadi · Sina Paryani · Aminreza Neshat · Ataollah Shirzadi · Shamsuddin Shahid · Ronak Ghanbari · Farzad Malakyar · John J. Clague) joint work.  This research focuses on enhancing the accuracy and efficiency of wetland mapping through the integration of #semi-supervised learning techniques and temporal analysis. The results indicate a notable increase in classification accuracy compared to traditional supervised learning methods. The study reveals important trends in the #Zrebar Wetland, such as a decrease in water bodies and fluctuations in reed areas from 1984 to 2022.  Findings highlight the changing nature of the #wetland ecosystem and underscore the importance of ongoing monitoring. more details here: (https://lnkd.in/gSgJUMnU) I would like to express my deepest appreciation to Dr. Hemin Shahabi and Dr. Mehdi Gholamnia for the opportunity to contribute to this significant research. Their guidance and support were invaluable throughout the process.

28

6 Comments

Investing in the future of our guardians :) .. this is huge! Many years and thoughts have been discussed since "Out of Our Minds" by Sir Ken Robinson, who simply alerted us that our pedagogic, economic, cultural and social well-being is contingent on our creative adaptations to the volatile dynamics. By the time we get those young generations at the University level, we will spend many years in damage control and RE-institutions of our own creative confidence. But what has been hammered down for many years as a system cannot be undone by asking them to think outside of the "box", having spent a lifetime fitting their conventions to the dimensions of the same damn box! @FUTUREGUARRDIAN.IO is envisioned to address this and much more, seeding for and nurturing local and sustainable talents. The investment made here for the guardians of our futures is only the first step on a 1,000-mile journey. Limited Seats. Apply Now https://lnkd.in/dZn6SH-x SAJORY Baynna Pupilar UAE | Future of Hiring

4

Research Topic: Autonomous Vehicles and Advancements in Object Detection Technology Dr. Mohammad ElYabroudi is one of our new faculty members who has already embarked on collecting data for his current research on autonomous vehicles. One of his research projects involves advancements in object detection technology, focusing specifically on traffic signs in the USA under challenging conditions. This research addresses complex scenarios such as severe weather conditions, shadows, and sun glare to improve the recognition of traffic signs, which are crucial for understanding traffic laws and directions. The initial phase involves extensive data collection, labeling, and preprocessing. The upcoming phase will see the release of a comprehensive public dataset uniquely centered on USA traffic signs. This dataset features meticulous pixel-level labels and includes various environmental attributes, such as: *** Weather Conditions: Sunny, foggy, snowy, and rainy *** Types of Roads: Highway, suburban *** Contextual Factors: Construction zones, school zones, residential zones *** Complexity Levels: Easy, moderate, complex These detailed attributes, combined with improved object detection algorithms, promise to significantly enhance the accuracy and reliability of traffic sign detection systems. Dr. ElYabroudi’s research gate profile: https://lnkd.in/enQeR8Jt #autonomous #ADAS #artificialintelligence #AI Lawrence Technological University Lawrence Tech Electrical and Computer Engineering (ECE)

24

2 Comments

Our very own PhD Students, Mahsa G. and Radhakrishnan Venkatakrishnan under the leadership of M Abdullah Canbaz, Ph.D. presented their research paper on the "Exploration of Large Language Models' Emotion Detection Abilities: Use Cases from the Middle East" at the IEEE Conference on Artificial Intelligence where they showcased their research findings on the potential of pre-trained transformer-based language models such as GTP 3.5 and RoBERTa for emotion detection in natural language processing, specifically focusing on the quality of emotional detection in LLMs. Findings include the emotional response of two significant events in Iran, Turkey and Syria. #aritificialintelligence #IEEE #ConferenceonArtificialIntelligence #LLMs

108

5 Comments

I had a blast talking about "The Power of (Modern) Convex Optimization in Nonconvex Perception, Control, and Learning" at Northeastern University. I highlighted two recent strands of research from my group: (1) Convex relaxation + large-scale numerical optimization + structure of robotics problems + new computing hardware such as GPUs + data, will provide a "reimagined" solution to nonconvex optimization problems that were previously considered too difficult to solve (e.g., structure from motion, trajectory optimization) (2) Adaptive optimization algorithms, in particular those recently introduced in the online convex optimization community, have surprising potential in designing "truly adaptive" optimizers that make learning/finetuning more stable, robust, efficient, and work "out of the box" (e.g., tackling loss of plasticity in lifelong reinforcement learning). Check out my slides if interested, and stay tuned for more that's coming! https://t.co/i8Ae2QXY5N

208

1 Comment

Thank you to Prof. Akramul Azim, PhD, PEng for giving a guest lecture today in my “Applications of LLMs to Software Engineering” grad course. He talked about his #CASCON2024 paper “FaultLLAMA2: A Fine Tuned Large Language Model for Code Recommendations to Reduce Java Lock Contention.” #OntarioTech #concurrency #LLM

20

Big update for students pursuing particle physics in Pakistan. PINSTECH is developing a 20 MeV linear electron accelerator, to be put into operation by July 2025. Snippet from a presentation given at the 6th international ion beam applications workshop at National Center of Physics (NCP).

72

1 Comment

📢 Excited to share recent research achievements with my PhD student Reza Vafaee Here are three of our latest papers: 1- Learning-based Sparse Sensing with Performance Guarantees in IEEE Transactions on Automatic Control A novel method for converting weighted sensor schedules to sparse ones while maintaining performance. Read more: https://lnkd.in/e4eD2QGF 2- Real-time Sensor Selection for Time-Varying Networks with Guaranteed Performance in Automatica Dynamic sensor selection for LTV systems with high performance and minimal sensors. Read more: https://lnkd.in/eCYTmb9c 3- Exploring Non-Submodular Scheduling for Large-Scale Sensor Networks in IEEE Control Systems Letters An effective greedy algorithm for non-submodular sensor scheduling. Read more: https://lnkd.in/etxfgJaS Proud of our progress in sparse sensing in complex networks! 🚀 #Research #SensorNetworks #ControlSystems #SparseSensing #IEEE #Automatica #PhDResearch

64

4 Comments

I’m happy to share that my PhD student, Sadia Saeed, has published an insightful article, "Review on the Role of Virtual Reality in Reducing Mental Health Diseases Specifically Stress, Anxiety, and Depression," on arXiv. This work highlights the potential of Virtual Reality (VR) as a transformative tool in mental health care, offering new ways to address conditions like stress, anxiety, and depression. I would like to express my gratitude to Abdul Qayyum, PhD from the National Heart and Lung Institute, Imperial College London, for his valuable guidance and support in this research. To access the preprint arXiv: https://lnkd.in/dQ_UN_np The Islamia University of Bahawalpur Pakistan Imperial College London #MentalHealth #VirtualReality #Research #Innovation #PhD

33

Robotics is a rapidly growing field in modern technologies. Check out the highlights of my Robotic team in the attached video! 🤖 [Link to the video]: https://lnkd.in/eQchj7sV

2

Congratulations to the authors, and especially to Mohadeseh Fallah, PhD student under my supervision at the 2nd year of the PhD in Systems Science, Track in Computational Mechanics, at Scuola IMT Alti Studi Lucca for her first article published in a prestigious international journal of mechanics of materials. The article is the result of an intensive work with novel micromechanical tests in the lab and a statistical mechanics approach to understand the behavior of paper and fibrous cellulose materials. The main results is the ability of the model to decouple the material constitutive relation of the fibers from the topological parameters of the fiber network: the material anisotropy is found to be the result of the different orientation of the population of fibers with respect to the direction of loading. The article contributes to the knowledge of complex fibrous materials and may have also an impact on technology transfer to the industrial district of Lucca, a worldwide excellence on paper and paperboard production. https://lnkd.in/d4G__Wm4

53

4 Comments