Safety-aware Motion Prediction with Unseen Vehicles for Autonomous Driving

Ren, Xuanchi; Yang, Tao; Li, Li Erran; Alahi, Alexandre; Chen, Qifeng

doi:10.1109/iccv48922.2021.01544

Cited by 18 publications

(6 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is convenient to limit the study to static occlusions because the evolution of the field of view scenario only depends on the location of the AV, which can be estimated if high-resolution map is available. Recent works have tackled the dynamic occlusion, such as [7][8][9][10], although each of them considered only a specific traffic scenario.…”

Section: Related Wordmentioning

confidence: 99%

Occlusion-Aware Motion Planning for Autonomous Driving

Wang

Zhou

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Motion planning of autonomous vehicles is still challenging in urban road environments with occlusions. In this work, with the view of safety, comfort and efficiency, we present a motion planning framework that enables autonomous vehicles to navigate safely in urban road with occlusions. Our solution mainly includes three parts: local path planning, trajectory planning and speed planning. First, based on the improved Artificial Potential Field to generate the local path, then the optimal trajectory is solved in the S-L coordinate with the local path as the reference line. Finally, the potential risk probability of the occluded area is incorporated into the incomplete information static game framework and implement speed planning based on the game results and the proposed vehicle "safe driving" to complete the collision avoidance between the autonomous vehicle and visible or obscured dynamic traffic participants. Through simulation verification in common traffic scenarios and a comparison with some existing methods, the proposed method is proven to enable autonomous vehicles to navigate safely in traffic scenarios with occlusions and improve the driving efficiency and comfort of autonomous vehicles.

show abstract

Section: Related Wordmentioning

confidence: 99%

Occlusion-Aware Motion Planning for Autonomous Driving

Wang

Zhou

et al. 2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The work of [24], [25] built the navigation map from visual inputs to learn the control policy. More recently, deep learning is widely applied to solve various problems in autonomous driving such as 3D object detection, building HD maps, and obstacle avoidance [26]- [30]. The authors in [31] investigated how the ground plane contributes to 3D detection in driving scenarios.…”

Section: Related Workmentioning

confidence: 99%

Deep Federated Learning for Autonomous Driving

Nguyen¹,

Do²,

Tran³

et al. 2021

Preprint

View full text Add to dashboard Cite

Autonomous driving is an active research topic in both academia and industry. However, most of the existing solutions focus on improving the accuracy by training learnable models with centralized large-scale data. Therefore, these methods do not take into account the user's privacy. In this paper, we present a new approach to learn autonomous driving policy while respecting privacy concerns. We propose a peer-to-peer Deep Federated Learning (DFL) approach to train deep architectures in a fully decentralized manner and remove the need for central orchestration. We design a new Federated Autonomous Driving network (FADNet) that can improve the model stability, ensure convergence, and handle imbalanced data distribution problems while is being trained with federated learning methods. Intensively experimental results on three datasets show that our approach with FADNet and DFL achieves superior accuracy compared with other recent methods. Furthermore, our approach can maintain privacy by not collecting user data to a central server. Our source code and trained models are available at https://github.com/aioz-ai/FADNet

show abstract

“…Human motion prediction is a fundamental problem in computer vision and has been broadly studied for decades. Modeling human motion can facilitate various kinds of applications in fields such as computer animation [12], [38] and antonymous driving [35], [34]. To solve the inherent ambiguity within human motion, recent work [4], [43], [29], [42], [44] focuses on modeling multimodal human motions, instead of the traditional deterministic regression for prediction [31], [23], [30], [28].…”

Section: Introductionmentioning

confidence: 99%

Learning Disentangled Representations for Controllable Human Motion Prediction

Gu¹,

Wang²,

Zhang³

2022

Preprint

View full text Add to dashboard Cite

Generative model-based motion prediction techniques have recently realized predicting controlled human motions, such as predicting multiple upper human body motions with similar lower-body motions. However, to achieve this, the state-of-the-art methods require either subsequently learning mapping functions to seek similar motions or training the model repetitively to enable control over the desired portion of body. In this paper, we propose a novel framework to learn disentangled representations for controllable human motion prediction. Our network involves a conditional variational auto-encoder (CVAE) architecture to model full-body human motion, and an extra CVAE path to learn only the corresponding partial-body (e.g., lower-body) motion. Specifically, the inductive bias imposed by the extra CVAE path encourages two latent variables in two paths to respectively govern separate representations for each partial-body motion. With a single training, our model is able to provide two types of controls for the generated human motions: (i) strictly controlling one portion of human body and (ii) adaptively controlling the other portion, by sampling from a pair of latent spaces. Additionally, we extend and adapt a sampling strategy to our trained model to diversify the controllable predictions. Our framework also potentially allows new forms of control by flexibly customizing the input for the extra CVAE path. Extensive experimental results and ablation studies demonstrate that our approach is capable of predicting state-of-the-art controllable human motions both qualitatively and quantitatively.

show abstract

Safety-aware Motion Prediction with Unseen Vehicles for Autonomous Driving

Cited by 18 publications

References 34 publications

Occlusion-Aware Motion Planning for Autonomous Driving

Occlusion-Aware Motion Planning for Autonomous Driving

Deep Federated Learning for Autonomous Driving

Learning Disentangled Representations for Controllable Human Motion Prediction

Contact Info

Product

Resources

About