Yebo Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Yebo Wu has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Li Li
University of Macau, Macau, Macao
7 shared publications - Cheng-Zhong Xu
State Key Laboratory of IoTSC, University of Macau, Macau, Macao
4 shared publications - Kahou Tam
University of Macau, Macau, Macao
4 shared publications - Chunlin Tian
State Key Laboratory of Internet of Things for Smart City, University of Macau, Macau, China
3 shared publications - Erwu Liu
College of Electronics and Information Engineering, Tongji University, Shanghai, China
2 shared publications - Jie Wang
Sino-German College of Intelligent Manufacturing, Shenzhen Technology University
2 shared publications
Latest Publications
- Open Access
- 17 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 144-157
Memory-Decoupled Layer-Wise Fine-Tuning for Efficient On-Device LLM Adaptation
- 17 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 144-157
Open AccessMemory-Decoupled Layer-Wise Fine-Tuning for Efficient On-Device LLM Adaptation
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 37 (11) , 2342-2358
Transcending Memory Constraints in Federated Learning via Sequential Block-Wise Training
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 37 (11) , 2342-2358
Transcending Memory Constraints in Federated Learning via Sequential Block-Wise Training
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4463-4474
HeimdaLLM: Efficient Cloud-assisted Federated Fine-tuning with Zeroth-Order Rectification for LLMs
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4463-4474
Open AccessHeimdaLLM: Efficient Cloud-assisted Federated Fine-tuning with Zeroth-Order Rectification for LLMs
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 37 (7) , 1630-1646
Floe : Federated Specialization for Real-Time LLM–SLM Inference
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 37 (7) , 1630-1646
Floe : Federated Specialization for Real-Time LLM–SLM Inference
- Open Access
Floe: Federated Specialization for Real-Time LLM-SLM Inference
Deploying large language models (LLMs) in real-time systems remains challenging due to their substantial computational demands and privacy concerns. We propose Floe, a hybrid federated learning framework designed for latency-sensitive, resource-constrained environments. Floe combines a cloud-based black-box LLM with lightweight small language models (SLMs) on edge devices to enable low-latency, privacy-preserving inference. Personal data and fine-tuning remain on-device, while the cloud LLM contributes general knowledge without exposing proprietary weights. A heterogeneity-aware LoRA adaptation strategy enables efficient edge deployment across diverse hardware, and a logit-level fusion mechanism enables real-time coordination between edge and cloud models. Extensive experiments demonstrate that Floe enhances user privacy and personalization. Moreover, it significantly improves model performance and reduces inference latency on edge devices under real-time constraints compared with baseline approaches.
Open AccessFloe: Federated Specialization for Real-Time LLM-SLM Inference
Deploying large language models (LLMs) in real-time systems remains challenging due to their substantial computational demands and privacy concerns. We propose Floe, a hybrid federated learning framework designed for latency-sensitive, resource-constrained environments. Floe combines a cloud-based black-box LLM with lightweight small language models (SLMs) on edge devices to enable low-latency, privacy-preserving inference. Personal data and fine-tuning remain on-device, while the cloud LLM contributes general knowledge without exposing proprietary weights. A heterogeneity-aware LoRA adaptation strategy enables efficient edge deployment across diverse hardware, and a logit-level fusion mechanism enables real-time coordination between edge and cloud models. Extensive experiments demonstrate that Floe enhances user privacy and personalization. Moreover, it significantly improves model performance and reduces inference latency on edge devices under real-time constraints compared with baseline approaches.
- Open Access
- 20 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1623-1632
Breaking the Memory Wall for Heterogeneous Federated Learning via Progressive Training
- 20 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1623-1632
Open AccessBreaking the Memory Wall for Heterogeneous Federated Learning via Progressive Training
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
Indoor Localization Fusing Inertial Navigation With Monocular Depth Estimation in Federated Learning Framework With Data Heterogeneity
In the field of indoor positioning, both user privacy and positioning accuracy are of paramount importance. Using mobile phones for indoor positioning is a simple, effective, and cost-efficient method; however, it poses privacy concerns for users and often fails to meet the precision required for most navigation needs. Therefore, within a heterogeneous federated learning (FL) framework, we propose a method that utilizes smartphone inertial measurement unit (IMU) data to obtain 2-D vector velocity, which is then further integrated to obtain indoor relative positions. Additionally, we estimate user distance based on commonly available indoor surveillance cameras data. The user receives the distance data forwarded by the edge server and implements fusion localization, which overcomes the limitations of IMU-based positioning. The proposed positioning system, named federated IMU monocular ranging location (FedIMRLoc), is validated through dataset simulation and real-world indoor environment experiments, demonstrating its superiority. Compared to FedAVG, central learning, and FedAVG +ultra-wideband (UWB), FedIMRLoc exhibited an average improvement in localization accuracy of 64.7%, 30.2%, and 27%.
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
Indoor Localization Fusing Inertial Navigation With Monocular Depth Estimation in Federated Learning Framework With Data Heterogeneity
In the field of indoor positioning, both user privacy and positioning accuracy are of paramount importance. Using mobile phones for indoor positioning is a simple, effective, and cost-efficient method; however, it poses privacy concerns for users and often fails to meet the precision required for most navigation needs. Therefore, within a heterogeneous federated learning (FL) framework, we propose a method that utilizes smartphone inertial measurement unit (IMU) data to obtain 2-D vector velocity, which is then further integrated to obtain indoor relative positions. Additionally, we estimate user distance based on commonly available indoor surveillance cameras data. The user receives the distance data forwarded by the edge server and implements fusion localization, which overcomes the limitations of IMU-based positioning. The proposed positioning system, named federated IMU monocular ranging location (FedIMRLoc), is validated through dataset simulation and real-world indoor environment experiments, demonstrating its superiority. Compared to FedAVG, central learning, and FedAVG +ultra-wideband (UWB), FedIMRLoc exhibited an average improvement in localization accuracy of 64.7%, 30.2%, and 27%.
- 14 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 35 (12) , 2513-2526
Breaking the Memory Wall for Heterogeneous Federated Learning via Model Splitting
Federated Learning (FL) enables multiple devices to collaboratively train a shared model while preserving data privacy. Ever-increasing model complexity coupled with limited memory resources on the participating devices severely bottlenecks the deployment of FL in real-world scenarios. Thus, a framework that can effectively break the memory wall while jointly taking into account the hardware and statistical heterogeneity in FL is urgently required. In this paper, we propose SmartSplit , a framework that effectively reduces the memory footprint on the device side while guaranteeing the training progress and model accuracy for heterogeneous FL through model splitting. Towards this end, SmartSplit employs a hierarchical structure to adaptively guide the overall training process. In each training round, the central manager, hosted on the server, dynamically selects the participating devices and sets the cutting layer by jointly considering the memory budget, training capacity, and data distribution of each device. The MEC manager, deployed within the edge server, proceeds to split the local model and perform training of the server-side portion. Meanwhile, it fine-tunes the splitting points based on the time-evolving statistical importance. The on-device manager, embedded inside each mobile device, continuously monitors the local training status while employing cost-aware checkpointing to match the runtime dynamic memory budget. Extensive experiments on representative datasets are conducted on both commercial off-the-shelf mobile device testbeds. The experimental results show that SmartSplit excels in FL training on highly memory-constrained mobile SoCs, offering up to a 94% peak latency reduction and 100-fold memory savings. It enhances accuracy performance by 1.49%-57.18% and adaptively adjusts to dynamic memory budgets through cost-aware recomputation
- 14 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 35 (12) , 2513-2526
Breaking the Memory Wall for Heterogeneous Federated Learning via Model Splitting
Federated Learning (FL) enables multiple devices to collaboratively train a shared model while preserving data privacy. Ever-increasing model complexity coupled with limited memory resources on the participating devices severely bottlenecks the deployment of FL in real-world scenarios. Thus, a framework that can effectively break the memory wall while jointly taking into account the hardware and statistical heterogeneity in FL is urgently required. In this paper, we propose SmartSplit , a framework that effectively reduces the memory footprint on the device side while guaranteeing the training progress and model accuracy for heterogeneous FL through model splitting. Towards this end, SmartSplit employs a hierarchical structure to adaptively guide the overall training process. In each training round, the central manager, hosted on the server, dynamically selects the participating devices and sets the cutting layer by jointly considering the memory budget, training capacity, and data distribution of each device. The MEC manager, deployed within the edge server, proceeds to split the local model and perform training of the server-side portion. Meanwhile, it fine-tunes the splitting points based on the time-evolving statistical importance. The on-device manager, embedded inside each mobile device, continuously monitors the local training status while employing cost-aware checkpointing to match the runtime dynamic memory budget. Extensive experiments on representative datasets are conducted on both commercial off-the-shelf mobile device testbeds. The experimental results show that SmartSplit excels in FL training on highly memory-constrained mobile SoCs, offering up to a 94% peak latency reduction and 100-fold memory savings. It enhances accuracy performance by 1.49%-57.18% and adaptively adjusts to dynamic memory budgets through cost-aware recomputation
- 24 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (2) , 3653-3661
FedINS2: A Federated-Edge-Learning-Based Inertial Navigation System With Segment Fusion
Modern inertial measurement units (IMUs) with low cost, small size, and low power consumption are the key to improving indoor positioning accuracy. However, the performance of IMU in current mobile phones is spotty. The IMU positioning error of mainstream cell phones is basically greater than 1%, which makes it difficult for indoor fusion positioning technology to achieve accuracy within 3 m on a mobile phone, seriously restricting the development of the industry. Therefore, the existing indoor fusion algorithms need to introduce additional hardware, such as high-density UWB/Wi-Fi/Bluetooth deployment, to compensate for the lack of IMU performance, which significantly increases the deployment cost, difficulty, and reduces environmental applicability. It is found that deep learning can greatly improve the performance of IMU, but traditional deep learning methods have problems such as privacy protection, difficulty in data collection, and low training efficiency. Therefore, we propose a novel data-driven inertial navigation method based on federated learning, named FedINS, to improve IMU performance and solve the above-mentioned problems. In order to further improve performance and reduce hardware cost, we introduce the concept of segment fusion. FedINS2 (2 means the second S, which is the segment) is formed by combining FedINS with low-cost edge-end ranging, which has terminal computing capabilities and edge-end ranging capabilities. FedINS2 not only greatly improves the performance of the smartphone IMU from 3.6% to 0.8%, but also has the characteristics of privacy protection and efficient data collection. Experimental results demonstrate that the proposed data-driven inertial navigation algorithm is effective.
- 24 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (2) , 3653-3661
FedINS2: A Federated-Edge-Learning-Based Inertial Navigation System With Segment Fusion
Modern inertial measurement units (IMUs) with low cost, small size, and low power consumption are the key to improving indoor positioning accuracy. However, the performance of IMU in current mobile phones is spotty. The IMU positioning error of mainstream cell phones is basically greater than 1%, which makes it difficult for indoor fusion positioning technology to achieve accuracy within 3 m on a mobile phone, seriously restricting the development of the industry. Therefore, the existing indoor fusion algorithms need to introduce additional hardware, such as high-density UWB/Wi-Fi/Bluetooth deployment, to compensate for the lack of IMU performance, which significantly increases the deployment cost, difficulty, and reduces environmental applicability. It is found that deep learning can greatly improve the performance of IMU, but traditional deep learning methods have problems such as privacy protection, difficulty in data collection, and low training efficiency. Therefore, we propose a novel data-driven inertial navigation method based on federated learning, named FedINS, to improve IMU performance and solve the above-mentioned problems. In order to further improve performance and reduce hardware cost, we introduce the concept of segment fusion. FedINS2 (2 means the second S, which is the segment) is formed by combining FedINS with low-cost edge-end ranging, which has terminal computing capabilities and edge-end ranging capabilities. FedINS2 not only greatly improves the performance of the smartphone IMU from 3.6% to 0.8%, but also has the characteristics of privacy protection and efficient data collection. Experimental results demonstrate that the proposed data-driven inertial navigation algorithm is effective.