Cheng Bi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Cheng Bi has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Ganchao Liu
School of Artificial Intelligence, OPtics and ElectroNics (iOPEN), Northwestern Polytechnical University, Xian, China
4 shared publications Jingfeng Wang
2 shared publicationsYuan Yuan
2 shared publications- Bangyong Sun
Faculty of Printing Packaging Engineering and Digital Media Technology Xi'an University of Technology Xi'an China
1 shared publication - Bingliang Hu
Key Laboratory of Spectral Imaging Technology of CAS, Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China
1 shared publication Bo Sun
School of Information Engineering, Changan University, Xian, China
1 shared publication
Latest Publications
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (13) , 29555-29571
From Depth to Saturation: Rethinking Small Model Capacity for Vehicle Detection in Resource-Constrained Environments
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (13) , 29555-29571
From Depth to Saturation: Rethinking Small Model Capacity for Vehicle Detection in Resource-Constrained Environments
- 15 May 2026
- Published by Elsevier in ISPRS Journal of Photogrammetry and Remote Sensing
- Vol. 238, 243-260
UAV coarse visual localization in large-scale continuous scenes
- 15 May 2026
- Published by Elsevier in ISPRS Journal of Photogrammetry and Remote Sensing
- Vol. 238, 243-260
UAV coarse visual localization in large-scale continuous scenes
- Open Access
- 26 December 2025
- Published by MDPI AG in Remote Sensing
- Vol. 18 (1) , 88
Beyond Spatial Domain: Multi-View Geo-Localization with Frequency-Based Positive-Incentive Information Screening
The substantial domain discrepancy inherent in multi-source and multi-view imagery presents formidable challenges to achieving precise drone-based multi-view geo-localization. Existing methodologies primarily focus on designing sophisticated backbone architectures to extract view-invariant representations within abstract feature spaces, yet they often overlook the rich and discriminative frequency-domain cues embedded in multi-view data. Inspired by the principles of -Noise theory, this paper proposes a frequency-domain Positive-Incentive Information Screening (PIIS) mechanism that adaptively identifies and preserves task-relevant frequency components based on entropy-guided information metrics. This principled approach selectively enhances discriminative spectral signatures while suppressing redundant or noisy components, thereby improving multi-view feature alignment under substantial appearance and geometric variations. The proposed PIIS strategy demonstrates strong architectural generality, as it can be seamlessly integrated into various backbone networks including convolutional-based and Transformer-based architectures while maintaining consistent performance improvements across different models. Extensive evaluations on the University-1652 and SUES-200 datasets have validated the great potential of the proposed method. Specifically, the PIIS-N model achieves a Recall@1 of 94.56% and a mean Average Precision (mAP) of 95.44% on the University-1652 dataset, exhibiting competitive accuracy among contemporary approaches. These findings underscore the considerable promise of frequency-domain analysis in advancing multi-view geo-localization.- 26 December 2025
- Published by MDPI AG in Remote Sensing
- Vol. 18 (1) , 88
Open AccessBeyond Spatial Domain: Multi-View Geo-Localization with Frequency-Based Positive-Incentive Information Screening
The substantial domain discrepancy inherent in multi-source and multi-view imagery presents formidable challenges to achieving precise drone-based multi-view geo-localization. Existing methodologies primarily focus on designing sophisticated backbone architectures to extract view-invariant representations within abstract feature spaces, yet they often overlook the rich and discriminative frequency-domain cues embedded in multi-view data. Inspired by the principles of -Noise theory, this paper proposes a frequency-domain Positive-Incentive Information Screening (PIIS) mechanism that adaptively identifies and preserves task-relevant frequency components based on entropy-guided information metrics. This principled approach selectively enhances discriminative spectral signatures while suppressing redundant or noisy components, thereby improving multi-view feature alignment under substantial appearance and geometric variations. The proposed PIIS strategy demonstrates strong architectural generality, as it can be seamlessly integrated into various backbone networks including convolutional-based and Transformer-based architectures while maintaining consistent performance improvements across different models. Extensive evaluations on the University-1652 and SUES-200 datasets have validated the great potential of the proposed method. Specifically, the PIIS-N model achieves a Recall@1 of 94.56% and a mean Average Precision (mAP) of 95.44% on the University-1652 dataset, exhibiting competitive accuracy among contemporary approaches. These findings underscore the considerable promise of frequency-domain analysis in advancing multi-view geo-localization. - 11 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (5) , 5968-5982
BEMN: Balanced Bias Enhanced Multi-Branch Network for Cross-View Geo-Localization
Cross-view geo-localization (CVGL) offers a promising alternative for positioning in GNSS-constrained environments through visual matching techniques. Extreme viewpoint variations and the complexity of real-world scenes present significant challenges to this task. However, current methods primarily focus on learning single-scale features, which may be inadequate for practical applications. Although some approaches attempt to incorporate multi-scale representations, they may suffer from unimodal bias arising from structural discrepancies among model branches, limiting effective multi-scale feature extraction. To address these issues, we propose a fully multi-branch network architecture, named BEMN, which is designed to learn multi-scale robust feature representations. Specifically, we construct a multi-branch backbone network based on pretrained visual models and design a two-stage training strategy. In the first stage, a separate training scheme is employed to thoroughly optimize each branch of the network, and a joint feature alignment (JFA) module is introduced to align cross-view features. The entire network is fine-tuned in the second stage, where a frequency domain adjustment (FDA) module is designed to improve performance. To further assess the generalization ability of CVGL methods, we establish Xian-37, a highly challenging CVGL test dataset featuring complex real scenes captured from diverse platforms and viewpoints. Experimental results across multiple public benchmarks validate the superiority of our approach, achieving state-of-the-art performance and demonstrating outstanding generalization capabilities. Our code and model are available at https://github.com/VERYBC/BEMN.
- 11 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (5) , 5968-5982
BEMN: Balanced Bias Enhanced Multi-Branch Network for Cross-View Geo-Localization
Cross-view geo-localization (CVGL) offers a promising alternative for positioning in GNSS-constrained environments through visual matching techniques. Extreme viewpoint variations and the complexity of real-world scenes present significant challenges to this task. However, current methods primarily focus on learning single-scale features, which may be inadequate for practical applications. Although some approaches attempt to incorporate multi-scale representations, they may suffer from unimodal bias arising from structural discrepancies among model branches, limiting effective multi-scale feature extraction. To address these issues, we propose a fully multi-branch network architecture, named BEMN, which is designed to learn multi-scale robust feature representations. Specifically, we construct a multi-branch backbone network based on pretrained visual models and design a two-stage training strategy. In the first stage, a separate training scheme is employed to thoroughly optimize each branch of the network, and a joint feature alignment (JFA) module is introduced to align cross-view features. The entire network is fine-tuned in the second stage, where a frequency domain adjustment (FDA) module is designed to improve performance. To further assess the generalization ability of CVGL methods, we establish Xian-37, a highly challenging CVGL test dataset featuring complex real scenes captured from diverse platforms and viewpoints. Experimental results across multiple public benchmarks validate the superiority of our approach, achieving state-of-the-art performance and demonstrating outstanding generalization capabilities. Our code and model are available at https://github.com/VERYBC/BEMN.
- 11 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (2) , 3982-3997
A Dynamic Fault Diagnosis Method for Gravitational Reference Sensor Based on Informer-RS
Gravitational Reference Sensor (GRS) used in the space gravitational wave detection mission may encounter intermittent failures characterized by random data jumps during its operation, potentially disrupting the overall mission progress. This paper first analyzes the pose sensing principle of the GRS and provides the differential capacitance fault model. In response to this type of fault, a data-driven dynamic fault diagnosis approach based on a novel Informer model is proposed. In this design, a random slice mechanism is introduced into the Informer model to enhance the robustness, and an anomaly-correcting temporal prediction method is employed to mitigate the influence of abnormal sequences in prediction. Then, considering the fault characteristics of the GRS, an improved isolation forest algorithm is proposed to calculate anomaly scores for detected sequences, and dynamic thresholds are established based on these scores for fault diagnosis. Finally, comparative experiments based on actual engineering data are conducted to verify the prediction performance of the method, and a series of simulation experiments based on Monte Carlo statistical analysis are conducted to evaluate the fault diagnosis capability of the proposed approach. The results indicate that the fault detection accuracy of this approach can reach 97.83%, and it outperforms some common approaches in terms of diagnostic performance and robustness.
- 11 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (2) , 3982-3997
A Dynamic Fault Diagnosis Method for Gravitational Reference Sensor Based on Informer-RS
Gravitational Reference Sensor (GRS) used in the space gravitational wave detection mission may encounter intermittent failures characterized by random data jumps during its operation, potentially disrupting the overall mission progress. This paper first analyzes the pose sensing principle of the GRS and provides the differential capacitance fault model. In response to this type of fault, a data-driven dynamic fault diagnosis approach based on a novel Informer model is proposed. In this design, a random slice mechanism is introduced into the Informer model to enhance the robustness, and an anomaly-correcting temporal prediction method is employed to mitigate the influence of abnormal sequences in prediction. Then, considering the fault characteristics of the GRS, an improved isolation forest algorithm is proposed to calculate anomaly scores for detected sequences, and dynamic thresholds are established based on these scores for fault diagnosis. Finally, comparative experiments based on actual engineering data are conducted to verify the prediction performance of the method, and a series of simulation experiments based on Monte Carlo statistical analysis are conducted to evaluate the fault diagnosis capability of the proposed approach. The results indicate that the fault detection accuracy of this approach can reach 97.83%, and it outperforms some common approaches in terms of diagnostic performance and robustness.