Qi Tan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Qi Tan has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Ke Xu
Tsinghua University, Beijing, China
2 shared publications - Beichen Li
School of Electrical and Information Engineering, Tianjin University, Tianjin, China
1 shared publication Dan Wang
Hong Kong Polytechnic University
1 shared publication- Guangmeng Zhou
Department of Computer Science and Technology, Tsinghua University, Beijing, China
1 shared publication Haiyang Wang
University of Minnesota at Duluth
1 shared publicationMeng Shen
Beijing Institute of Technology
1 shared publication
Latest Publications
- 12 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Networking
- Vol. 32 (3) , 1873-1887
FedPAGE: Pruning Adaptively Toward Global Efficiency of Heterogeneous Federated Learning
When workers are heterogeneous in computing and transmission capabilities, the global efficiency of federated learning suffers from the straggler issue, i.e., the slowest worker drags down the overall training process. We propose a novel and efficient federated learning framework named FedPAGE, where workers perform distributed pruning adaptively towards global efficiency, i.e., fast training and high accuracy. For fast training, we develop a pruning rate learning approach generating an adaptive pruning rate for each worker, making the overall update time approximate to the fastest worker’s update time, i.e., no stragglers. For high accuracy, we find that structural similarity between sub-models is essential to global model accuracy in the distributed pruning, and thus propose the CIG_X pruning scheme to ensure maximum similarity. Meanwhile, we adopt the sparse training and design model aggregating of different size sub-models to cope with distributed pruning. We prove the convergence of FedPAGE and demonstrate the effectiveness of FedPAGE on image classification and natural language inference tasks. Compared with the state-of-the-art, FedPAGE achieves higher accuracy with the same speedup ratio.
- 12 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Networking
- Vol. 32 (3) , 1873-1887
FedPAGE: Pruning Adaptively Toward Global Efficiency of Heterogeneous Federated Learning
When workers are heterogeneous in computing and transmission capabilities, the global efficiency of federated learning suffers from the straggler issue, i.e., the slowest worker drags down the overall training process. We propose a novel and efficient federated learning framework named FedPAGE, where workers perform distributed pruning adaptively towards global efficiency, i.e., fast training and high accuracy. For fast training, we develop a pruning rate learning approach generating an adaptive pruning rate for each worker, making the overall update time approximate to the fastest worker’s update time, i.e., no stragglers. For high accuracy, we find that structural similarity between sub-models is essential to global model accuracy in the distributed pruning, and thus propose the CIG_X pruning scheme to ensure maximum similarity. Meanwhile, we adopt the sparse training and design model aggregating of different size sub-models to cope with distributed pruning. We prove the convergence of FedPAGE and demonstrate the effectiveness of FedPAGE on image classification and natural language inference tasks. Compared with the state-of-the-art, FedPAGE achieves higher accuracy with the same speedup ratio.
- 2 December 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 72 (00189456) , 1-12
GAN-Based Radar Spectrogram Augmentation via Diversity Injection Strategy
The classification of human activity using radar has gained considerable attention in recent years because of the radar sensor’s resistance to harsh settings. However, when using machine learning algorithms to train a radar-based classifier, a substantial amount of training data is always required, which means that the time-consuming and diffcult radar measurements should be a prerequisite activity in most scenarios. The purpose of this paper is to present a novel data augmentation methodology for generating suffcient and diverse training data for human activity classification, hence alleviating the reliance on complex radar measurements. It is worth noting that the proposed model only uses one spectrogram as a reference for spectrogram augmentation, which demonstrates its great potential in practical scenarios. Considering the contingent risk of a lack of diversity in augmented samples, we develop an elaborate strategy for injecting diversity into augmented samples using external counterparts. To validate our model, we conduct radar simulations and measurements to create a variety of datasets. Our model outperforms other comparable models, demonstrating its great potential in improving the performance of human activity classification.
- 2 December 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 72 (00189456) , 1-12
GAN-Based Radar Spectrogram Augmentation via Diversity Injection Strategy
The classification of human activity using radar has gained considerable attention in recent years because of the radar sensor’s resistance to harsh settings. However, when using machine learning algorithms to train a radar-based classifier, a substantial amount of training data is always required, which means that the time-consuming and diffcult radar measurements should be a prerequisite activity in most scenarios. The purpose of this paper is to present a novel data augmentation methodology for generating suffcient and diverse training data for human activity classification, hence alleviating the reliance on complex radar measurements. It is worth noting that the proposed model only uses one spectrogram as a reference for spectrogram augmentation, which demonstrates its great potential in practical scenarios. Considering the contingent risk of a lack of diversity in augmented samples, we develop an elaborate strategy for injecting diversity into augmented samples using external counterparts. To validate our model, we conduct radar simulations and measurements to create a variety of datasets. Our model outperforms other comparable models, demonstrating its great potential in improving the performance of human activity classification.
- 29 November 2019
- Published by Springer Nature in CCF Transactions on Networking
- Vol. 2 (3-4) , 188-206
Interference-aware lifetime maximization with joint routing and charging in wireless sensor networks
- 29 November 2019
- Published by Springer Nature in CCF Transactions on Networking
- Vol. 2 (3-4) , 188-206
Interference-aware lifetime maximization with joint routing and charging in wireless sensor networks