Jie Jia
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jie Jia has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Bingzhao Gao
Tongji University
1 shared publication - Chenmin Ye
School of Information Engineering, Shanghai Maritime University, Shanghai 201306, P. R. China
1 shared publication - Chunjiang Fan
No. 8 Wuxi People’s Hospital
1 shared publication - Jinyang Zhuang
Department of Rehabilitation Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200025, China
1 shared publication - Lei Cao
College of Information Engineering, Shanghai Maritime University, Shanghai, China
1 shared publication - Wenchao Ding
Fudan University
1 shared publication
Latest Publications
- 19 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (4) , 4457-4464
Learning a Unified Risk Map for Autonomous Driving in Partially Observable Environments
Occlusion-aware prediction remains a critical challenge in autonomous driving due to the inherent uncertainty of unobserved regions. Existing approaches either overestimate risk based on reachable states or struggle to predict accurate trajectories under high occlusion uncertainty. To address these limitations, we propose a unified risk map modeling and learning framework for partially observable environments. Our method integrates traffic flow risk and collision risk through spatiotemporal modeling, enabling fine-grained assessment of occlusion-induced hazards. To address the scarcity of scenarios involving occluded interactions, we introduce a diffusion-based scenario generation framework that produces realistic yet adversarial scenarios. We integrate the modeling and learning of a unified risk map into a framework that supports risk-aware planning under partial observability. Experiments on the Waymo Open Motion Dataset show that our method significantly outperforms the state-of-the-art occlusion-aware baseline, improving minimum time-to-collision by 0.78 times and average time-to-collision by 1.67 times. The proposed framework offers a comprehensive and practical solution for risk-aware planning in partially observable environments.
- 19 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (4) , 4457-4464
Learning a Unified Risk Map for Autonomous Driving in Partially Observable Environments
Occlusion-aware prediction remains a critical challenge in autonomous driving due to the inherent uncertainty of unobserved regions. Existing approaches either overestimate risk based on reachable states or struggle to predict accurate trajectories under high occlusion uncertainty. To address these limitations, we propose a unified risk map modeling and learning framework for partially observable environments. Our method integrates traffic flow risk and collision risk through spatiotemporal modeling, enabling fine-grained assessment of occlusion-induced hazards. To address the scarcity of scenarios involving occluded interactions, we introduce a diffusion-based scenario generation framework that produces realistic yet adversarial scenarios. We integrate the modeling and learning of a unified risk map into a framework that supports risk-aware planning under partial observability. Experiments on the Waymo Open Motion Dataset show that our method significantly outperforms the state-of-the-art occlusion-aware baseline, improving minimum time-to-collision by 0.78 times and average time-to-collision by 1.67 times. The proposed framework offers a comprehensive and practical solution for risk-aware planning in partially observable environments.
- 7 August 2025
- Published by World Scientific Pub Co Pte Ltd in Journal of Mechanics in Medicine and Biology
- Vol. 25 (08)
A NOVEL INTEGRATION OF MULTI-SCALE AND TEMPORAL CONVOLUTION ALGORITHM FOR ENHANCING STROKE REHABILITATION THROUGH AN INNOVATIVE COMBINATION OF MIRRORED ACTION OBSERVATION AND MOTOR IMAGERY
Brain–computer interface (BCI) techniques, especially those based on electroencephalogram (EEG), have demonstrated substantial value in stroke rehabilitation by enhancing neuroplasticity and motor skills. This study seeks to further improve stroke rehabilitation by leveraging brain plasticity through the combination of mirrored action observation and motor imagery (CMAOMI). To tackle classification challenges in stroke patients, a novel multi-scale convolutional-attention temporal network (MCAT) is introduced. MCAT integrates multi-scale convolutional neural networks (CNNs) and temporal convolutional networks (TCNs) to capture temporal features across various frequency domains. By incorporating the Squeeze-and-Excitation (SE) module and TCN layers, MCAT refines temporal information, thereby enhancing classification accuracy. The experimental results demonstrate that CMAOMI exhibits significantly better EEG classification accuracy than both motor imagery (MI) and mirrored action observation (MAO). Additionally, MCAT achieves the highest average accuracy of 88.03%, outperforming existing baseline models. This study provides a novel approach for stroke rehabilitation by combining advanced neuroscience and artificial intelligence technology, which plays a pivotal role in the creation of effective BCI training protocols for stroke rehabilitation.
- 7 August 2025
- Published by World Scientific Pub Co Pte Ltd in Journal of Mechanics in Medicine and Biology
- Vol. 25 (08)
A NOVEL INTEGRATION OF MULTI-SCALE AND TEMPORAL CONVOLUTION ALGORITHM FOR ENHANCING STROKE REHABILITATION THROUGH AN INNOVATIVE COMBINATION OF MIRRORED ACTION OBSERVATION AND MOTOR IMAGERY
Brain–computer interface (BCI) techniques, especially those based on electroencephalogram (EEG), have demonstrated substantial value in stroke rehabilitation by enhancing neuroplasticity and motor skills. This study seeks to further improve stroke rehabilitation by leveraging brain plasticity through the combination of mirrored action observation and motor imagery (CMAOMI). To tackle classification challenges in stroke patients, a novel multi-scale convolutional-attention temporal network (MCAT) is introduced. MCAT integrates multi-scale convolutional neural networks (CNNs) and temporal convolutional networks (TCNs) to capture temporal features across various frequency domains. By incorporating the Squeeze-and-Excitation (SE) module and TCN layers, MCAT refines temporal information, thereby enhancing classification accuracy. The experimental results demonstrate that CMAOMI exhibits significantly better EEG classification accuracy than both motor imagery (MI) and mirrored action observation (MAO). Additionally, MCAT achieves the highest average accuracy of 88.03%, outperforming existing baseline models. This study provides a novel approach for stroke rehabilitation by combining advanced neuroscience and artificial intelligence technology, which plays a pivotal role in the creation of effective BCI training protocols for stroke rehabilitation.