A5108057302- Vehicle License Plate Recognition
- Speech Recognition and Synthesis
- Advanced Neural Network Applications
- Anomaly Detection Techniques and Applications
- Autonomous Vehicle Technology and Safety
- Nematode management and characterization studies
- Dental Radiography and Imaging
- Natural Language Processing Techniques
- Fault Detection and Control Systems
- Adversarial Robustness in Machine Learning
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Gaze Tracking and Assistive Technology
- Handwritten Text Recognition Techniques
- Biosimilars and Bioanalytical Methods
- Medical Imaging and Analysis
- Microbial Metabolic Engineering and Bioproduction
- Topic Modeling
- Entomopathogenic Microorganisms in Pest Control
- Gait Recognition and Analysis
- Bluetooth and Wireless Communication Technologies
- Wireless Networks and Protocols
- Hand Gesture Recognition Systems
- Time Series Analysis and Forecasting
- Augmented Reality Applications
- VLSI and FPGA Design Techniques
East China University of Science and Technology
2024-2025
Swinburne University of Technology
2025
University of Essex
2024
National University of Computer and Emerging Sciences
2011-2022
Case Western Reserve University
2022
University School
2022
Korea Advanced Institute of Science and Technology
2021
Technical University of Munich
2021
Deutsches Zentrum für Luft- und Raumfahrt e. V. (DLR)
2020-2021
Heinrich Heine University Düsseldorf
2018-2020
Due to their increasing spread, confidence in neural network predictions became more and important. However, basic networks do not deliver certainty estimates or suffer from over under confidence. Many researchers have been working on understanding quantifying uncertainty a network's prediction. As result, different types sources of identified variety approaches measure quantify proposed. This work gives comprehensive overview estimation networks, reviews recent advances the field,...
Deep learning in remote sensing has received considerable international hype, but it is mostly limited to the evaluation of optical data. Although deep been introduced synthetic aperture radar (SAR) data processing, despite successful first attempts, its huge potential remains locked. In this article, we provide an introduction most relevant models and concepts, point out possible pitfalls by analyzing special characteristics SAR data, review state art applied SAR, summarize available...
Abstract Background Adipose-derived stem cells (ADSCs) have been extensively used in preclinical and clinical trials for treating various diseases. However, the differences between ADSCs from lean individuals (L-ADSCs) those obese (O-ADSCs) not thoroughly investigated, particularly regarding their mitochondrial lysosomal functions. Therefore, this study aims to evaluate L-ADSCs O-ADSCs terms of cell biological activity, mitochondria, lysosomes. Methods We first isolated cultured O-ADSCs....
ABSTRACT High‐performance strain and corresponding fermentation process are essential for achieving efficient biomanufacturing. However, conventional offline detection methods products cumbersome less stable, hindering the “Test” module in operation of “Design‐Build‐Test‐Learn” cycle screening optimization. This study proposed validated an innovative research paradigm combining computer vision with deep learning to facilitate selection effective A practical framework was developed gentamicin...
As electric vehicles (EVs) become more integrated into intelligent transportation systems, vast amounts of personal and operational data, such as location, driving patterns, energy consumption, are continuously collected. Ensuring privacy for this sensitive data is critical to prevent tracking, profiling, unauthorised access. This paper presents the implementation event-wise differential (DP) safeguard individual points in EV ecosystems, focusing on protecting event-level information like...
Vehicle make and model recognition plays an important role in monitoring traffic a vehicle surveillance system. Identifying is challenging task due to intraclass variation, view-point different illumination conditions (Hassan et al., 2021). In this domain, many datasets regarding car e.g. Stanford Car (Krause 2013), VMMRdB (Tafazzoli 2017, Yang 2015), have already been experimented with by researchers. However, most of the images these are high-quality no conditions. Further, collected...
Abstract Artificial Intelligence (AI) technology is spearheading a new industrial revolution, which provides ample opportunities for the transformational development of traditional fermentation processes. During plasmid fermentation, subjective process control leads to highly unstable yields. In this study, multi‐parameter correlation analysis was first performed discover dynamic metabolic balance among oxygen uptake rate, temperature, and yield, whilst revealing heating rate timing as most...
Vehicle make and model recognition is an important component of the Intelligent Transport System (ITS), which plays essential role in vehicle surveillance traffic monitoring. In this paper, we evaluated performance recent deep neural networks for to identify 196 different types vehicles based on their make, model, year using Stanford Cars data. Transfer learning has been employed reduce training time also added a dropout 0.5 at last dense layer before output improved generalization applied...
Interferometric Synthetic Aperture Radar (InSAR)-derived surface displacement time series enable a wide range of applications from urban structural monitoring to geohazard assessment. With systematic data acquisitions becoming the new norm for SAR missions, millions are continuously generated. Machine Learning provides framework efficient mining such big data. Here, we focus on unsupervised via clustering similar temporal patterns and data-driven signal reconstruction InSAR series. We...
As the use of biomedical signals is incredibly increasing in both clinical and nonclinical applications. They have a great deal development devices that can be controlled by information inferred from thoughts. One current hot topics for research Brain Computer Interface (BCI) on basis EEG signals. BCI technology makes humans to control computer or other BCIs given new hopes people who suffer locked-in syndrome motor disabilities providing alternative means communication channels. The...
Deep Learning is often criticized as being a black-box method that provides accurate predictions, but limited explanation of the underlying processes and no indication when to not trust those predictions. Equipping existing deep learning models with an (general) notion uncertainty can help mitigate both these issues. The Bayesian community has developed model-agnostic methodology estimate data model be implemented on top models. In this work, we test for recurrent satellite time series...
The matter of handwritten text recognition is as yet a major challenge to mainstream researchers.A few ways deal with this have been endeavored in the most recent years, for part concentrating on English pre-printed or characters space.Consequently, need effort research concerning Arabic texts recognition.The handwriting presents unique technical difficulties because it cursive, right left writing and letters convert its shapes structures when putted at initial, middle, isolation end...
This paper demonstrates a system that can translate American Sign Language into text from real time video with computer vision based approach. has two parts, Skin Color Detection (SCD) and Hand Gesture Recognition (HGR). pixels are detected the image by applying some heuristic rules using an Artificial Neural Network (ANN). The ANN is trained H, S, V component values HSV color space along five texture features. Automated threshold value for SCD achieved plotting Q YIQ vs result each pixel...
In this paper, we propose two algorithms for generating a complete n-bit binary reflected Gray code sequence. The first one is called Backtracking. It generates sequence by only sub-tree instead of the tree. approach, both and its reflection generated concatenating "0" "1" to most significant bit position n-1 result at each leaf node sub-tree. second MOptimal. modification Space time optimal approach [8] considering minimization total number outer inner loop execution purpose. method, also...
Simulations have gained paramount importance in terms of software development for wireless sensor networks and been a vital focus the scientific community this decade to provide efficient, secure, safe communication smart cities. Network Simulators are widely used secure architectures city. Therefore, technical survey report, we conducted experimental comparisons among ten different simulation environments that can be simulate smart-city operations. We comprehensively analyze compare...
Objectives: Age and gender estimation is crucial for various applications, including forensic investigations anthropological studies. This research aims to develop a predictive system age in living individuals, leveraging dental measurements such as Coronal Height (CH), Pulp Cavity (CPCH), Tooth Index (TCI). Methods: Machine learning models were employed our study, Cat Boost Classifier (Catboost), Gradient Boosting (GBM), Ada (AdaBoost), Random Forest (RF), eXtreme (XGB), Light (LGB), Extra...