Zheqi Yu

Dr. Zheqi Yu

University of Glasgow
orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 March 2023 in Remote Sensing
Verified
Journal Article
Open access
PDF file
QWQIAN WANG
QIAN WANG
WLWenfeng Li
Wenfeng Li
ZYZheqi Yu
Zheqi Yu
Qammer AbbasiQammer Abbasi
Qammer Abbasi
Muhammad ImranMuhammad Imran
Muhammad Imran
TZTong Zhu
Tong Zhu
In recent years, major natural disasters and public safety accidents have frequently occurred worldwide. In order to deal with various disasters and accidents using rapidly deployable, reliable, efficient, and stable emergency communication networks, all countries in the world are strengthening and improving emergency communication network construction and related technology research. Motivated by these situations, in this paper, we provide a state-of-the-art survey of the current situation and development of emergency communication networks. In this detailed investigation, our primary focus is the extensive discussion of emergency communication network technology, including satellite networks, ad hoc networks, cellular networks, and wireless private networks. Then, we explore and analyze the networks currently applied in emergency rescue, such as the 370M narrowband private network, broadband cluster network, and 5G constellation plan. We propose a broadband-narrowband integrated emergency communication network to provide an effective solution for visual dispatch of emergency rescue services. The main findings derived from the comprehensive survey on the emergency communication network are then summarized, and possible research challenges are noted. Lastly, we complete this survey by shedding new light on future directions for the emergency communication network. In the future, the emergency network will develop in the direction of intelligence, integration, popularization, and lower cost, and space-air-ground-sea integrated networks. This survey provides a reference basis for the construction of networks to mitigate major natural disasters and public safety accidents.

1 January 2020 in IEEE Access
Verified
Journal Article
Open access
ZYZheqi Yu
Zheqi Yu
AAAmir M. Abdulghani
Amir M. Abdulghani
AZAdnan Zahid
Adnan Zahid
Hadi HeidariHadi Heidari
Hadi Heidari
Muhammad ImranMuhammad Imran
Muhammad Imran
Qammer AbbasiQammer Abbasi
Qammer Abbasi
Compared with von Neumann's computer architecture, neuromorphic systems offer more unique and novel solutions to the artificial intelligence discipline. Inspired by biology, this novel system has implemented the theory of human brain modeling by connecting feigned neurons and synapses to reveal the new neuroscience concepts. Many researchers have vastly invested in neuro-inspired models, algorithms, learning approaches, operation systems for the exploration of the neuromorphic system and have implemented many corresponding applications. Recently, some researchers have demonstrated the capabilities of Hopfield algorithms in some large-scale notable hardware projects and seen significant progression. This paper presents a comprehensive review and focuses extensively on the Hopfield algorithm's model and its potential advancement in new research applications. Towards the end, we conclude with a broad discussion and a viable plan for the latest application prospects to facilitate developers with a better understanding of the aforementioned model in accordance to build their own artificial intelligence projects.

17 August 2022 in IEEE Sensors Journal
Verified
Journal Article
Open access
ZYZheqi Yu
Zheqi Yu
Ahmad TahaAhmad Taha
Ahmad Taha
WTWilliam Taylor
William Taylor
AZadnan zahid
adnan zahid
KRKhalid Rajab
Khalid Rajab
Qammer AbbasiQammer Abbasi
Qammer Abbasi
This article provides a new benchmark dataset for 3-D point cloud classification in which the manually labeled human activity data exceeds 100 point clouds per frame and is capable of meeting the training needs for data-intensive learning approaches. In this study, a case study is considered for evaluating the benchmark using a deep long short-term memory (LSTM) neural network, which demonstrated a significant performance improvement over the state-of-the-art human activity recognition (HAR) area. To date, numerous types of collection devices have been used in the recognition of human activities. However, due to the scarcity of training data, the task of 3-D point cloud labeling has not yet made significant progress. To overcome this challenge, it is aimed to deduce this data requirements gap, allowing deep-learning methods to reach their full potential in 3-D point cloud tasks. The dataset used for this process is comprised of dense point clouds acquired with the static ground sensor by the NodeNs company-supported multiple input multiple output (MIMO) radar (NodeNs ZERO 60 GHz IQ radar). It contains multiple types of human being data ranging from one to four individuals and encompasses a range of human action scenarios, including standing, sitting, picking up, falling, and walking. Furthermore, it also investigated sensor locations and requirements for human being data collection that is from a single subject to multiple subjects, as well as identified and analyzed various sensing devices and applications that collect activity data. In this regard, a thorough study is conducted on several benchmark datasets, examining sensors, characteristics, activity categories, and other data. Finally, it compares and analyzes the activity recognition methods used in several benchmark datasets based on the current study. Unlike existing devices, the new NodeNs sensor provides more accessible and straightforward point cloud data to capture human movement information. Depending on an advanced detection algorithm to process point cloud data, it achieved more than 95% accuracy on the benchmark dataset.

2 November 2020 in Electronics
Verified
Journal Article
Open access
PDF file
ZYZheqi Yu
Zheqi Yu
Pedro MachadoPedro Machado
Pedro Machado
AZadnan zahid
adnan zahid
AAAmir M. Abdulghani
Amir M. Abdulghani
KDKia Dashtipour
Kia Dashtipour
Qammer AbbasiQammer Abbasi
Qammer Abbasi
This paper suggests an optimisation approach in heterogeneous computing systems to balance energy power consumption and efficiency. The work proposes a power measurement utility for a reinforcement learning (PMU-RL) algorithm to dynamically adjust the resource utilisation of heterogeneous platforms in order to minimise power consumption. A reinforcement learning (RL) technique is applied to analyse and optimise the resource utilisation of field programmable gate array (FPGA) control state capabilities, which is built for a simulation environment with a Xilinx ZYNQ multi-processor systems-on-chip (MPSoC) board. In this study, the balance operation mode for improving power consumption and performance is established to dynamically change the programmable logic (PL) end work state. It is based on an RL algorithm that can quickly discover the optimization effect of PL on different workloads to improve energy efficiency. The results demonstrate a substantial reduction of 18% in energy consumption without affecting the application’s performance. Thus, the proposed PMU-RL technique has the potential to be considered for other heterogeneous computing platforms.

6 September 2022 in IEEE Internet of Things Journal
Verified
Journal Article
Open access
ZYZheqi Yu
Zheqi Yu
AZadnan zahid
adnan zahid
Ahmad TahaAhmad Taha
Ahmad Taha
WTWilliam Taylor
William Taylor
Julien Le KernecJulien Le Kernec
Julien Le Kernec
Qammer AbbasiQammer Abbasi
Qammer Abbasi
The increasing demand for considering multisensor data fusion technology has drawn attention for precise human activity recognition (HAR) over standalone technology due to its reliability and robustness. This article presents a framework that fuses data from multiple sensing systems and applies neuromorphic computing to sense and classify human activities. The data is collected by utilizing inertial measurement unit (IMU) sensors, software-defined radios, and radars, and feature extraction and selection are performed on the data. For each of the actions, such as sitting and standing, an activity matrix is generated, which is then fed into a discrete Hopfield neural network as a binary feature pattern for one-shot learning. Following the Hopfield network neurons’ feedback output, the conformity to the standard activity feature pattern is also determined. Following the Hopfield network neurons’ feedback output, the training of neurons is completed after two steps under the Hebbian learning law, and the conformity to the standard activity feature pattern is also determined. According to the probabilistic statistics on inference predictions, the proposed method, that is the neuromorphic computing of the three data fused framework, achieved the box plot for the highest lower quartile output of 95.34%, while the confusion matrix classification accuracy of the two activities was 98.98%. The results have shown that neuromorphic computing is most capable of multisensor data-fusion-based HAR. Furthermore, the proposed method can be enhanced by incorporating additional hardware signal processing in the system to enable the flexible integration of human activity data.

17 December 2020 in Sensors
Verified
Journal Article
Open access
PDF file
ZYZheqi Yu
Zheqi Yu
AZadnan zahid
adnan zahid
SAShuja Ansari
Shuja Ansari
HAhasan tahir abbas
hasan tahir abbas
AAAmir M. Abdulghani
Amir M. Abdulghani
Qammer AbbasiQammer Abbasi
Qammer Abbasi
With the popularity of smart wearable systems, sensor signal processing poses more challenges to machine learning in embedded scenarios. For example, traditional machine-learning methods for data classification, especially in real time, are computationally intensive. The deployment of Artificial Intelligence algorithms on embedded hardware for fast data classification and accurate fall detection poses a huge challenge in achieving power-efficient embedded systems. Therefore, by exploiting the associative memory feature of Hopfield Neural Network, a hardware module has been designed to simulate the Neural Network algorithm which uses sensor data integration and data classification for recognizing the fall. By adopting the Hebbian learning method for training neural networks, weights of human activity features are obtained and implemented/embedded into the hardware design. Here, the neural network weight of fall activity is achieved through data preprocessing, and then the weight is mapped to the amplification factor setting in the hardware. The designs are checked with validation scenarios, and the experiment is completed with a Hopfield neural network in the analog module. Through simulations, the classification accuracy of the fall data reached 88.9% which compares well with some other results achieved by the software-based machine-learning algorithms, which verify the feasibility of our hardware design. The designed system performs the complex signal calculations of the hardware’s feedback signal, replacing the software-based method. A straightforward circuit design is used to meet the weight setting from the Hopfield neural network, which is maximizing the reusability and flexibility of the circuit design.