Continual Pedestrian Trajectory Learning With Social Generative Replay

Wu, Ya; Bighashdel, Ariyan; Chen, Guang; Dubbelman, Gijs; Jancura, Pavol

doi:10.1109/lra.2022.3231833

Cited by 11 publications

(3 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fortunately, the combination of the IoT and AI has resulted in significant advancements in ITS [222]- [224]. These technologies enable interconnectivity among various traffic participants, including vehicles, traffic signals, humans, and infrastructure, to enhance transportation efficiency, reduce emissions, and prevent accidents.…”

Section: Collaborative Control Of Cavsmentioning

confidence: 99%

A Systematic Survey of Control Techniques and Applications in Connected and Automated Vehicles

Liu,

Hua,

Deng

et al. 2023

IEEE Internet Things J.

View full text Add to dashboard Cite

Vehicle control is one of the most critical challenges in autonomous vehicles (AVs) and connected and automated vehicles (CAVs), and it is paramount in vehicle safety, passenger comfort, transportation efficiency, and energy saving. This survey attempts to provide a comprehensive and thorough overview of the current state of vehicle control technology, focusing on the evolution from vehicle state estimation and trajectory tracking control in AVs at the microscopic level to collaborative control in CAVs at the macroscopic level. First, this review starts with vehicle key state estimation, specifically vehicle sideslip angle, which is the most pivotal state for vehicle trajectory control, to discuss representative approaches. Then, we present symbolic vehicle trajectory tracking control approaches for AVs. On top of that, we further review the collaborative control frameworks for CAVs and corresponding applications. Finally, this survey concludes with a discussion of future research directions and the challenges. This survey aims to provide a contextualized and indepth look at the state of the art in vehicle control for AVs and CAVs, identifying critical areas of focus and pointing out the potential areas for further exploration.

show abstract

Section: Collaborative Control Of Cavsmentioning

confidence: 99%

A Systematic Survey of Control Techniques and Applications in Connected and Automated Vehicles

Liu,

Hua,

Deng

et al. 2023

IEEE Internet Things J.

View full text Add to dashboard Cite

show abstract

“…Social LSTM [2] models the trajectories of individual agents from separate LSTM networks and aggregates the LSTM hidden cues to model their interactions. CL-SGR [42] considers the sample replay model in a continuous trajectory prediction scenario setting to avoid catastrophic forgetting. The other branch [17,31,43] models the interaction among the agents based on the attention mechanism.…”

Section: Related Workmentioning

confidence: 99%

IPCC-TP: Utilizing Incremental Pearson Correlation Coefficient for Joint Multi-Agent Trajectory Prediction

Zhu¹,

Zhai²,

Yan³

et al. 2023

Preprint

View full text Add to dashboard Cite

Reliable multi-agent trajectory prediction is crucial for the safe planning and control of autonomous systems. Compared with single-agent cases, the major challenge in simultaneously processing multiple agents lies in modeling complex social interactions caused by various driving intentions and road conditions. Previous methods typically leverage graph-based message propagation or attention mechanism to encapsulate such interactions in the format of marginal probabilistic distributions. However, it is inherently sub-optimal. In this paper, we propose IPCC-TP, a novel relevance-aware module based on Incremental Pearson Correlation Coefficient to improve multi-agent interaction modeling. IPCC-TP learns pairwise joint Gaussian Distributions through the tightly-coupled estimation of the means and covariances according to interactive incremental movements. Our module can be conveniently embedded into existing multi-agent prediction methods to extend original motion distribution decoders. Extensive experiments on nuScenes and Argoverse 2 datasets demonstrate that IPCC-TP improves the performance of baselines by a large margin.

show abstract

“…Social LSTM (Alahi et al, 2016) models the trajectories of individual agents from separate LSTM networks and aggregates the LSTM hidden cues to model their interactions. CL-SGR (Wu et al, 2022) considers the sample replay model in a continuous trajectory prediction scenario setting to avoid catastrophic forgetting. The other branch (Girgis et al, 2021) models the interaction among the agents based on the attention mechanism.…”

Section: Multi-agent Trajectory Modelingmentioning

confidence: 99%

Res-FLNet: human-robot interaction and collaboration for multi-modal sensing robot autonomous driving tasks based on learning control algorithm

Wang

2023

Front. Neurorobot.

View full text Add to dashboard Cite

IntroductionRes-FLNet presents a cutting-edge solution for addressing autonomous driving tasks in the context of multimodal sensing robots while ensuring privacy protection through Federated Learning (FL). The rapid advancement of autonomous vehicles and robotics has escalated the need for efficient and safe navigation algorithms that also support Human-Robot Interaction and Collaboration. However, the integration of data from diverse sensors like cameras, LiDARs, and radars raises concerns about privacy and data security.MethodsIn this paper, we introduce Res-FLNet, which harnesses the power of ResNet-50 and LSTM models to achieve robust and privacy-preserving autonomous driving. The ResNet-50 model effectively extracts features from visual input, while LSTM captures sequential dependencies in the multimodal data, enabling more sophisticated learning control algorithms. To tackle privacy issues, we employ Federated Learning, enabling model training to be conducted locally on individual robots without sharing raw data. By aggregating model updates from different robots, the central server learns from collective knowledge while preserving data privacy. Res-FLNet can also facilitate Human-Robot Interaction and Collaboration as it allows robots to share knowledge while preserving privacy.Results and discussionOur experiments demonstrate the efficacy and privacy preservation of Res-FLNet across four widely-used autonomous driving datasets: KITTI, Waymo Open Dataset, ApolloScape, and BDD100K. Res-FLNet outperforms state-of-the-art methods in terms of accuracy, robustness, and privacy preservation. Moreover, it exhibits promising adaptability and generalization across various autonomous driving scenarios, showcasing its potential for multi-modal sensing robots in complex and dynamic environments.

show abstract

Continual Pedestrian Trajectory Learning With Social Generative Replay

Cited by 11 publications

References 28 publications

A Systematic Survey of Control Techniques and Applications in Connected and Automated Vehicles

A Systematic Survey of Control Techniques and Applications in Connected and Automated Vehicles

IPCC-TP: Utilizing Incremental Pearson Correlation Coefficient for Joint Multi-Agent Trajectory Prediction

Res-FLNet: human-robot interaction and collaboration for multi-modal sensing robot autonomous driving tasks based on learning control algorithm

Contact Info

Product

Resources

About