PRECOG: PREdiction Conditioned On Goals in Visual Multi-Agent Settings

Rhinehart, Nicholas; McAllister, Rowan; Levine, Sergey

doi:10.48550/arxiv.1905.01296

Cited by 20 publications

(65 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 3 presents the evaluation results against available benchmarks. Our model outperforms most of these prediction models [4,6,10,22,23], except MFP [16]. Although MFP achieved the best performance, MFP is a deterministic model that outputs a fixed number of deterministic modes (i.e., trajectories), compared to other methods including ours that outputs a non-deterministic trajectories.…”

Section: Quantitative Resultsmentioning

confidence: 93%

“…MinMSD is a measure of diversity or multi-modality of the model outputs. It is often used for evaluating self-driving prediction models, in particular, generative prediction models that construct probabilistic models with sampling capability [4,16,22,23]. The definition of MinMSD is as follows.…”

Section: Quantitative Resultsmentioning

confidence: 99%

“…Normalizing flow, or Flow, is also a type of explicit generative model. While flow-based models [4,5,22] have access to the exact probability density p(x), flow-based models are typically better suited for density estimations than generative tasks (e.g., trajectory forecasting).…”

Section: Related Workmentioning

confidence: 99%

“…In this section, we demonstrate the performance of CVAE-H in urban driving scenarios using PRECOG-Carla dataset [22], which a publicly available dataset created using an open-source simulator called Carla [44].…”

Section: Experiments: Multi-agent Forecastingmentioning

confidence: 99%

See 3 more Smart Citations

CVAE-H: Conditionalizing Variational Autoencoders via Hypernetworks and Trajectory Forecasting for Autonomous Driving

Oh¹,

Peng²

2022

Preprint

View full text Add to dashboard Cite

The task of predicting stochastic behaviors of road agents in diverse environments is a challenging problem for autonomous driving. To best understand scene contexts and produce diverse possible future states of the road agents adaptively in different environments, a prediction model should be probabilistic, multi-modal, context-driven, and general. We present Conditionalizing Variational AutoEncoders via Hypernetworks (CVAE-H); a conditional VAE that extensively leverages hypernetwork and performs generative tasks for high-dimensional problems like the prediction task. We first evaluate CVAE-H on simple generative experiments to show that CVAE-H is probabilistic, multimodal, context-driven, and general. Then, we demonstrate that the proposed model effectively solves a self-driving prediction problem by producing accurate predictions of road agents in various environments.

show abstract

Section: Quantitative Resultsmentioning

confidence: 93%

Section: Quantitative Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Experiments: Multi-agent Forecastingmentioning

confidence: 99%

See 2 more Smart Citations

CVAE-H: Conditionalizing Variational Autoencoders via Hypernetworks and Trajectory Forecasting for Autonomous Driving

Oh¹,

Peng²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Other approaches also introduce structure into the task by exploring goal-conditioned ideas which model actor intent and goals prior to predicting the trajectory. Rhinehart et al [23] propose PRECOG, a goal-conditioned approach for multi-agent prediction. More recently, Mangalam et al [24,25] propose an endpoint conditioned prediction scheme which conditions pedestrian predictions on goal destinations.…”

Section: Related Workmentioning

confidence: 99%

Physically Feasible Vehicle Trajectory Prediction

Girase,

Hoang,

Yalamanchi

et al. 2021

Preprint

View full text Add to dashboard Cite

Predicting the future motion of actors in a traffic scene is a crucial part of any autonomous driving system. Recent research in this area has focused on trajectory prediction approaches that optimize standard trajectory error metrics. In this work, we describe three important properties -physical realism guarantees, system maintainability, and sample efficiency -which we believe are equally important for developing a self-driving system that can operate safely and practically in the real world. Furthermore, we introduce PTNet (PathTrackingNet), a novel approach for vehicle trajectory prediction that is a hybrid of the classical pure pursuit path tracking algorithm and modern graph-based neural networks. By combining a structured robotics technique with a flexible learning approach, we are able to produce a system that not only achieves the same level of performance as other state-of-the-art methods on traditional trajectory error metrics, but also provides strong guarantees about the physical realism of the predicted trajectories while requiring half the amount of data. We believe focusing on this new class of hybrid approaches is an useful direction for developing and maintaining a safety-critical autonomous driving system. * indicates equal contribution Preprint. Under review.

show abstract

Implicit Latent Variable Model for Scene-Consistent Motion Forecasting

Casas

Gulino²,

Suo

et al. 2020

Lecture Notes in Computer Science

100

108

View full text Add to dashboard Cite

In order to plan a safe maneuver an autonomous vehicle must accurately perceive its environment, and understand the interactions among traffic participants. In this paper, we aim to learn sceneconsistent motion forecasts of complex urban traffic directly from sensor data. In particular, we propose to characterize the joint distribution over future trajectories via an implicit latent variable model. We model the scene as an interaction graph and employ powerful graph neural networks to learn a distributed latent representation of the scene. Coupled with a deterministic decoder, we obtain trajectory samples that are consistent across traffic participants, achieving state-of-the-art results in motion forecasting and interaction understanding. Last but not least, we demonstrate that our motion forecasts result in safer and more comfortable motion planning.

show abstract

PRECOG: PREdiction Conditioned On Goals in Visual Multi-Agent Settings

Cited by 20 publications

References 0 publications

CVAE-H: Conditionalizing Variational Autoencoders via Hypernetworks and Trajectory Forecasting for Autonomous Driving

CVAE-H: Conditionalizing Variational Autoencoders via Hypernetworks and Trajectory Forecasting for Autonomous Driving

Physically Feasible Vehicle Trajectory Prediction

Implicit Latent Variable Model for Scene-Consistent Motion Forecasting

Contact Info

Product

Resources

About