A5025583139- Autonomous Vehicle Technology and Safety
- Caching and Content Delivery
- Advanced Data Storage Technologies
- Cooperative Communication and Network Coding
- Traffic and Road Safety
- Traffic Prediction and Management Techniques
- Vehicle emissions and performance
- Traffic control and management
- Cellular Automata and Applications
- Cryptography and Data Security
- Advanced Neural Network Applications
- Reinforcement Learning in Robotics
- Energy, Environment, and Transportation Policies
- Transportation and Mobility Innovations
- Smart Agriculture and AI
Shanghai Jiao Tong University
2018-2025
Tsinghua University
2018-2021
Probabilistic trajectory prediction for other vehicles can be an effective way to improve the understanding of dynamic and stochastic traffic environment automated vehicles. One challenge is how predict vehicle accurately both in short-term long-term horizon. In this paper, we propose integrated approach combining driver characteristic intention estimation (DCIE) model with Gaussian process (GP) model. Our proposed method makes use low-level high-level information inquires parameters by...
On a highway, the frequency of occurrence irrelevant features, such as trees, varies lot in different scenes. A limitation deep conventional neural networks used end-to-end self-driving systems is that if incoming images contain too much information, it makes difficult for network to extract only subset features required decision making. Consequently, while existing approaches may perform well training scenes, they not work correctly other In this study, we developed novel method an...
Purpose For autonomous vehicles, trajectory prediction of surrounding vehicles is beneficial to improving the situational awareness dynamic and stochastic traffic environments, which a crucial indispensable element realize highly automated driving. Design/methodology/approach In this paper, overall framework consists two parts: first, novel driver characteristic intention estimation (DCIE) model built indicate higher-level information vehicle using its low-level motion variables; then,...
The end-to-end control method is one of the ways to realize autonomous driving. Existing self-driving approaches commonly just consider objective, for instance, safety or fuel efficiency. To optimize multiple objectives simultaneously, Pareto-optimal driving policies need be obtained satisfy different requirements. However, it difficult obtain policy set because high-dimensional and complicated input in task. handle this problem, we propose actor-critic approach, which updating rule...
In distributed storage systems (DSSs), the optimal tradeoff between node and repair bandwidth is an important issue for designing coding strategies to ensure large scale data reliability. The capacity of DSSs obtained as a function parameters, characterizing tradeoff. There are lots works on with clusters (racks) where bandwidths from intra-cluster cross-cluster differentiated. However, separate nodes also prevalent in realistic DSSs, but (CSN-DSSs) insufficient. this paper, we formulate...