Yan Pan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yan Pan has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Bo Zhou
Southeast University
6 shared publications - Jiajie Wu
School of Automation, Southeast University and the Key Laboratory of Measurement and Control of CSE, Ministry of Education, Nanjing, China
3 shared publications - Chuanzhao Lu
School of Automation, Southeast University and the Key Laboratory of Measurement and Control of CSE, Ministry of Education, Nanjing, China
2 shared publications - Jiapeng Xie
School of Automation, Southeast University and the Key Laboratory of Measurement and Control of CSE, Ministry of Education, Nanjing, P. R. China
2 shared publications - Yueqi Zhu
School of Automation, Southeast University and the Key Laboratory of Measurement and Control of CSE, Ministry of Education, Nanjing, China
2 shared publications - Chufan Rui
School of Automation, Southeast University and the Key Laboratory of Measurement and Control of CSE, Ministry of Education, Nanjing, China
1 shared publication
Latest Publications
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (4) , 5214-5221
Sequential Probabilistic Descriptor via Uncertainty-Aware Multi-Modal Fusion for Safety-Critical Place Recognition
The success of loop closure and global map consistency largely relies on robust place recognition, making it a safety-critical capability for mobile robots in autonomous navigation. However, most existing approaches prioritize recognition accuracy while overlooking uncertainty estimation, which may lead to overconfident false predictions and compromise navigation safety. To address this limitation, this letter proposes a sequential multi-modal probabilistic descriptor for place representation, explicitly modeling uncertainty at both the modality and decision levels. Specifically, sequential image and LiDAR inputs leverage temporal information and are encoded as probabilistic embeddings, followed by an uncertainty-aware fusion. By adaptively balancing the contribution of each modality, prediction conflicts caused by sensor degradation are mitigated. Predictive uncertainty is further estimated through probabilistic combination rules, providing reliable confidence measures that allow downstream tasks to filter out high-uncertainty predictions. Extensive evaluations on NCLT, Oxford Robotcar and WildScenes datasets are conducted, covering long-term tasks, seasonal changes, illumination variations, and unstructured natural environments. Results show that the proposed method achieves superior performance across diverse conditions and provides well-calibrated uncertainty estimates that enable reliable failure detection, highlighting its potential for deployment in safety-critical scenes.
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (4) , 5214-5221
Sequential Probabilistic Descriptor via Uncertainty-Aware Multi-Modal Fusion for Safety-Critical Place Recognition
The success of loop closure and global map consistency largely relies on robust place recognition, making it a safety-critical capability for mobile robots in autonomous navigation. However, most existing approaches prioritize recognition accuracy while overlooking uncertainty estimation, which may lead to overconfident false predictions and compromise navigation safety. To address this limitation, this letter proposes a sequential multi-modal probabilistic descriptor for place representation, explicitly modeling uncertainty at both the modality and decision levels. Specifically, sequential image and LiDAR inputs leverage temporal information and are encoded as probabilistic embeddings, followed by an uncertainty-aware fusion. By adaptively balancing the contribution of each modality, prediction conflicts caused by sensor degradation are mitigated. Predictive uncertainty is further estimated through probabilistic combination rules, providing reliable confidence measures that allow downstream tasks to filter out high-uncertainty predictions. Extensive evaluations on NCLT, Oxford Robotcar and WildScenes datasets are conducted, covering long-term tasks, seasonal changes, illumination variations, and unstructured natural environments. Results show that the proposed method achieves superior performance across diverse conditions and provides well-calibrated uncertainty estimates that enable reliable failure detection, highlighting its potential for deployment in safety-critical scenes.
- 16 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (9) , 8826-8833
Safety-Critical Ultra-Wideband 3D Localization With Set-Membership Uncertainty Representation
Existing localization methods frequently focus on offline benchmarking based on datasets and assume that the methods possess the same performance in online applications without ground truth. However, since datasets always provide a limited number of scenarios, the reliability of localization methods in unknown scenarios cannot be inferred from their performance on the datasets. Therefore, the over-reliance on benchmarking results will pose a safety risk to the practical applications of localization methods. Inspired by this challenge, in this letter, we propose a safety-critical localization system that can measure the reliability of the estimated locations online in real time based on the set-membership uncertainty. By considering various system uncertainties and addressing the imbalance issue within the current set-membership filters, we design a dimension-balanced set-membership filter (DB-SMF) for ultra-wideband 3D localization based on the unknown but bounded (UBB) assumption and the dimension-balanced objective functions. Compared with the Gaussian uncertainty, our set-membership uncertainty can cover the unknown ground-truth locations with bounded sets. In safety-critical scenarios, this uncertainty can provide deterministic state bounds for tasks such as motion control and obstacle avoidance, thereby better ensuring the safety of the robot. Real-world experiments show that our DB-SMF can estimate set-membership uncertainties with the ability to cover unknown ground truth in finite spaces from range measurements and in this way ensure the safety of the localization system.
- 16 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (9) , 8826-8833
Safety-Critical Ultra-Wideband 3D Localization With Set-Membership Uncertainty Representation
Existing localization methods frequently focus on offline benchmarking based on datasets and assume that the methods possess the same performance in online applications without ground truth. However, since datasets always provide a limited number of scenarios, the reliability of localization methods in unknown scenarios cannot be inferred from their performance on the datasets. Therefore, the over-reliance on benchmarking results will pose a safety risk to the practical applications of localization methods. Inspired by this challenge, in this letter, we propose a safety-critical localization system that can measure the reliability of the estimated locations online in real time based on the set-membership uncertainty. By considering various system uncertainties and addressing the imbalance issue within the current set-membership filters, we design a dimension-balanced set-membership filter (DB-SMF) for ultra-wideband 3D localization based on the unknown but bounded (UBB) assumption and the dimension-balanced objective functions. Compared with the Gaussian uncertainty, our set-membership uncertainty can cover the unknown ground-truth locations with bounded sets. In safety-critical scenarios, this uncertainty can provide deterministic state bounds for tasks such as motion control and obstacle avoidance, thereby better ensuring the safety of the robot. Real-world experiments show that our DB-SMF can estimate set-membership uncertainties with the ability to cover unknown ground truth in finite spaces from range measurements and in this way ensure the safety of the localization system.
- 27 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63 (01962892) , 1-14
MAR-VO: A Match-and-Refine Framework for UAV’s Monocular Visual Odometry in Planetary Environments
In planetary exploration, unmanned aerial vehicles (UAVs) play a crucial role in visual navigation, yet face the challenges of GNSS absence and environment degeneration, such as low-texture and non-structural features. To address this issue, we first propose MAR (match and refine), a unified deep learning-based feature matching framework. In the matching stage, it incorporates a retentive self-attention (RSA) module to enhance feature matching accuracy. In the refining stage, local feature windows surrounding the key-points are fed into a transformer network to refine the feature point positions and filter out mismatches. Then we develop MAR-VO, a monocular visual odometry system leveraging MAR to localize in planetary environments using feature extractor such as SuperPoint. To test the effectiveness in such scenarios, we construct a virtual planet-like dataset named Terrain-UAV with 24 sequences including 22800 images in total, using models built in blender. Experimental results show that MAR achieves competitive results in homography estimation and relative pose estimation on benchmark datasets, while outperforms on Terrain-UAV. In addition, our MAR-VO system yields better absolute pose estimation RMSE than monocular point-based VO systems using traditional features and deep learning methods on Terrain-UAV. Notably, with MAR framework, our VO system exhibits excellent generality, especially in low-texture planetary scenes.
- 27 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63 (01962892) , 1-14
MAR-VO: A Match-and-Refine Framework for UAV’s Monocular Visual Odometry in Planetary Environments
In planetary exploration, unmanned aerial vehicles (UAVs) play a crucial role in visual navigation, yet face the challenges of GNSS absence and environment degeneration, such as low-texture and non-structural features. To address this issue, we first propose MAR (match and refine), a unified deep learning-based feature matching framework. In the matching stage, it incorporates a retentive self-attention (RSA) module to enhance feature matching accuracy. In the refining stage, local feature windows surrounding the key-points are fed into a transformer network to refine the feature point positions and filter out mismatches. Then we develop MAR-VO, a monocular visual odometry system leveraging MAR to localize in planetary environments using feature extractor such as SuperPoint. To test the effectiveness in such scenarios, we construct a virtual planet-like dataset named Terrain-UAV with 24 sequences including 22800 images in total, using models built in blender. Experimental results show that MAR achieves competitive results in homography estimation and relative pose estimation on benchmark datasets, while outperforms on Terrain-UAV. In addition, our MAR-VO system yields better absolute pose estimation RMSE than monocular point-based VO systems using traditional features and deep learning methods on Terrain-UAV. Notably, with MAR framework, our VO system exhibits excellent generality, especially in low-texture planetary scenes.
- 24 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (12) , 11162-11169
ATI-CTLO: Adaptive Temporal Interval-Based Continuous-Time LiDAR-Only Odometry
The motion distortion in LiDAR scans caused by the robot's aggressive motion and environmental terrain features significantly impacts the positioning and mapping performance of 3D LiDAR odometry. Existing distortion correction solutions struggle to balance computational complexity and accuracy. In this letter, we propose an A daptive T emporal I nterval-based C ontinuous- T ime L iDAR-only O dometry (ATI-CTLO), which is based on straightforward and efficient linear interpolation. Our method can flexibly adjust the temporal intervals between control nodes according to the motion dynamics and environmental degeneracy. This adaptability enhances performance across various motion states and improves the algorithms robustness in degenerate, particularly feature-sparse, environments. We validated our method's effectiveness on multiple datasets across different platforms, achieving comparable accuracy to state-of-the-art LiDAR-only odometry methods. Notably, in situations involving aggressive motion and sparse features, our method outperforms existing LiDAR-only methods.
- 24 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (12) , 11162-11169
ATI-CTLO: Adaptive Temporal Interval-Based Continuous-Time LiDAR-Only Odometry
The motion distortion in LiDAR scans caused by the robot's aggressive motion and environmental terrain features significantly impacts the positioning and mapping performance of 3D LiDAR odometry. Existing distortion correction solutions struggle to balance computational complexity and accuracy. In this letter, we propose an A daptive T emporal I nterval-based C ontinuous- T ime L iDAR-only O dometry (ATI-CTLO), which is based on straightforward and efficient linear interpolation. Our method can flexibly adjust the temporal intervals between control nodes according to the motion dynamics and environmental degeneracy. This adaptability enhances performance across various motion states and improves the algorithms robustness in degenerate, particularly feature-sparse, environments. We validated our method's effectiveness on multiple datasets across different platforms, achieving comparable accuracy to state-of-the-art LiDAR-only odometry methods. Notably, in situations involving aggressive motion and sparse features, our method outperforms existing LiDAR-only methods.
- 4 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. 72 (3) , 2801-2809
Camera-LiDAR Fusion With Latent Correlation for Cross-Scene Place Recognition
As a crucial issue for outdoor mobile robot navigation, place recognition (PR) remains challenging in long-term cross-scene applications, necessitating enhanced robustness of PR algorithms. In this article, a novel multimodal PR approach is proposed, which deeply fuses the camera and LiDAR to effectively compensate for their respective limitations. The introduced place descriptor consists of three branches: image-based, point cloud-based, and fusion-based. Specifically, the fusion-based branch employs a dual-stage pipeline, leveraging the latent correlation between the two modalities for information interaction and conducting channel-level fusion. Moreover, implicit alignment in the fusion branch, coupled with information supplementation from the two single-modal branches, better addresses the information loss caused by limited overlap in their field-of-view (FoV). In extensive experiments on Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI), North Campus Long-Term (NCLT), USVInland datasets, and the real-world campus test, the proposed method stands out in precision-recall curves with an average maximum F1 score of 0.949, demonstrating superior robustness and generalization compared with other state-of-the-art methods.
- 4 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. 72 (3) , 2801-2809
Camera-LiDAR Fusion With Latent Correlation for Cross-Scene Place Recognition
As a crucial issue for outdoor mobile robot navigation, place recognition (PR) remains challenging in long-term cross-scene applications, necessitating enhanced robustness of PR algorithms. In this article, a novel multimodal PR approach is proposed, which deeply fuses the camera and LiDAR to effectively compensate for their respective limitations. The introduced place descriptor consists of three branches: image-based, point cloud-based, and fusion-based. Specifically, the fusion-based branch employs a dual-stage pipeline, leveraging the latent correlation between the two modalities for information interaction and conducting channel-level fusion. Moreover, implicit alignment in the fusion branch, coupled with information supplementation from the two single-modal branches, better addresses the information loss caused by limited overlap in their field-of-view (FoV). In extensive experiments on Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI), North Campus Long-Term (NCLT), USVInland datasets, and the real-world campus test, the proposed method stands out in precision-recall curves with an average maximum F1 score of 0.949, demonstrating superior robustness and generalization compared with other state-of-the-art methods.
- 18 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 8 (12) , 8074-8081
MotionBEV: Attention-Aware Online LiDAR Moving Object Segmentation With Bird's Eye View Based Appearance and Motion Features
Identifying moving objects is an essential capability for autonomous systems, as it provides critical information for pose estimation, navigation, collision avoidance, and static map construction. In this letter, we present MotionBEV, a fast and accurate framework for LiDAR moving object segmentation, which segments moving objects with appearance and motion features in the bird's eye view (BEV) domain. Our approach converts 3D LiDAR scans into a 2D polar BEV representation to improve computational efficiency. Specifically, we learn appearance features with a simplified PointNet and compute motion features through the height differences of consecutive frames of point clouds projected onto vertical columns in the polar BEV coordinate system. We employ a dual-branch network bridged by the Appearance-Motion Co-attention Module (AMCM) to adaptively fuse the spatio-temporal information from appearance and motion features. Our approach achieves state-of-the-art performance on the SemanticKITTI-MOS benchmark. Furthermore, to demonstrate the practical effectiveness of our method, we provide a LiDAR-MOS dataset recorded by a solid-state LiDAR, which features non-repetitive scanning patterns and a small field of view.
- 18 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 8 (12) , 8074-8081
MotionBEV: Attention-Aware Online LiDAR Moving Object Segmentation With Bird's Eye View Based Appearance and Motion Features
Identifying moving objects is an essential capability for autonomous systems, as it provides critical information for pose estimation, navigation, collision avoidance, and static map construction. In this letter, we present MotionBEV, a fast and accurate framework for LiDAR moving object segmentation, which segments moving objects with appearance and motion features in the bird's eye view (BEV) domain. Our approach converts 3D LiDAR scans into a 2D polar BEV representation to improve computational efficiency. Specifically, we learn appearance features with a simplified PointNet and compute motion features through the height differences of consecutive frames of point clouds projected onto vertical columns in the polar BEV coordinate system. We employ a dual-branch network bridged by the Appearance-Motion Co-attention Module (AMCM) to adaptively fuse the spatio-temporal information from appearance and motion features. Our approach achieves state-of-the-art performance on the SemanticKITTI-MOS benchmark. Furthermore, to demonstrate the practical effectiveness of our method, we provide a LiDAR-MOS dataset recorded by a solid-state LiDAR, which features non-repetitive scanning patterns and a small field of view.