StarNet: Pedestrian Trajectory Prediction using Deep Neural Network in Star Topology

Zhu, Yuemin; Qian, Deheng; Ren, Dongchun; Xia, Huaxia

doi:10.48550/arxiv.1906.01797

Cited by 15 publications

(13 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…SGAN used the random Gaussian noise as the latent variable, and thus generated diverse outputs. Zhu et al [23] proposed StarNet, which is similar to SGAN except for the use of a query module. Amirian et al [24] replaced the L2 loss used in SGAN with the information loss [25] to avoid mode collapse.…”

Section: A Trajectory Prediction Methodsmentioning

confidence: 99%

TPPO: A Novel Trajectory Predictor with Pseudo Oracle

Yang¹,

Yan²,

Wang³

et al. 2020

Preprint

View full text Add to dashboard Cite

Forecasting pedestrian trajectories in dynamic scenes remains a critical problem with various applications, such as autonomous driving and socially aware robots. Such forecasting is challenging due to human-human and humanobject interactions and future uncertainties caused by human randomness. Generative model-based methods handle future uncertainties by sampling a latent variable. However, few previous studies carefully explored the generation of the latent variable. In this work, we propose the Trajectory Predictor with Pseudo Oracle (TPPO), which is a generative model-based trajectory predictor. The first pseudo oracle is pedestrians' moving directions, and the second one is the latent variable estimated from observed trajectories. A social attention module is used to aggregate neighbors' interactions on the basis of the correlation between pedestrians' moving directions and their future trajectories. This correlation is inspired by the fact that a pedestrian's future trajectory is often influenced by pedestrians in front. A latent variable predictor is proposed to estimate latent variable distributions from observed and ground-truth trajectories. Moreover, the gap between these two distributions is minimized during training. Therefore, the latent variable predictor can estimate the latent variable from observed trajectories to approximate that estimated from ground-truth trajectories. We compare the performance of TPPO with related methods on several public datasets. Results demonstrate that TPPO outperforms state-ofthe-art methods with low average and final displacement errors. Besides, the ablation study shows that the prediction performance will not dramatically decrease as sampling times decline during tests.

show abstract

Section: A Trajectory Prediction Methodsmentioning

confidence: 99%

TPPO: A Novel Trajectory Predictor with Pseudo Oracle

Yang¹,

Yan²,

Wang³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Amir et al [23] embedded two attention modules into the trajectory predictor for an improved prediction performance. Zhu et al [24] improved the SGAN by adding a query module. Learning information from the scene is another way to produce improved prediction performance.…”

Section: A Trajectory Prediction Methodsmentioning

confidence: 99%

A Novel Graph based Trajectory Predictor with Pseudo Oracle

Yang,

Yan,

Wang

et al. 2020

Preprint

View full text Add to dashboard Cite

Pedestrian trajectory prediction in dynamic scenes remains a challenging and critical problem in numerous applications, such as self-driving cars and socially aware robots. Challenges concentrate on capturing pedestrians' social interactions and handling their future uncertainties. Pedestrians' head orientations can be used as an oracle that indicates relevant pedestrians [1], thus is beneficial to model social interactions. Moreover, latent variable distributions of pedestrians' future trajectories can be termed as another oracle. However, few works fully utilize these oracle information for an improved prediction performance. In this work, we propose GTPPO (Graph-based Trajectory Predictor with Pseudo Oracle), which is a generative model-based trajectory predictor. Pedestrians' social interactions are captured by the proposed GA 2 T (Graph Attention & social Attention neTwork) module. Social attention is calculated on the basis of pedestrians' moving directions, which are termed as a pseudo oracle. Moreover, we propose a latent variable predictor to learn the latent variable distribution from observed trajectories. Such latent variable distribution reflects pedestrians' future trajectories, and therefore can be taken as another pseudo oracle. We compare the performance of GTPPO with several recently proposed methods on benchmarking datasets. Quantitative evaluations demonstrate that GTPPO outperforms state-of-theart methods with lower average and final displacement errors. Qualitative evaluations show that GTPPO successfully recognizes the sudden motion changes since the estimated latent variable reflects the future trajectories.

show abstract

“…Social GAN [1] introduces a loss that encourages the generative network of GAN to spread its distribution and cover the space of possible paths and designs a new pooling mechanism that is employed to encode subtle cues for all people involved in a scene. In [2], Zhu et al propose StarNet that has a star topology with both host network and hub one, where the hub network takes the observed trajectories of all pedestrians and produces a comprehensive spatio-temporal representation of all interactions in crowds. In order to integrate human social behaviors in dynamic scenes for prediction tasks, Sun et al propose a recursive social behavior graph [3], which is demonstrated to obtain greater expressive power and higher performance.…”

Section: B Time Series Prediction Based On Deep Learningmentioning

confidence: 99%

Phase Space Reconstruction Network for Lane Intrusion Action Recognition

Zhang¹,

Deng²,

Lin³

et al. 2021

Preprint

View full text Add to dashboard Cite

In a complex road traffic scene, illegal lane intrusion of pedestrians or cyclists constitutes one of the main safety challenges in autonomous driving application. In this paper, we propose a novel object-level phase space reconstruction network (PSRNet) for motion time series classification, aiming to recognize lane intrusion actions that occur 150m ahead through a monocular camera fixed on moving vehicle. In the PSRNet, the movement of pedestrians and cyclists, specifically viewed as an observable object-level dynamic process, can be reconstructed as trajectories of state vectors in a latent phase space and further characterized by a learnable Lyapunov exponent-like classifier that indicates discrimination in terms of average exponential divergence of state trajectories. Additionally, in order to first transform video inputs into one-dimensional motion time series of each object, a lane width normalization based on visual object tracking-by-detection is presented. Extensive experiments are conducted on the THU-IntrudBehavior dataset collected from real urban roads. The results show that our PSRNet could reach the best accuracy of 98.0%, which remarkably exceeds existing action recognition approaches by more than 30%.

show abstract

StarNet: Pedestrian Trajectory Prediction using Deep Neural Network in Star Topology

Cited by 15 publications

References 22 publications

TPPO: A Novel Trajectory Predictor with Pseudo Oracle

TPPO: A Novel Trajectory Predictor with Pseudo Oracle

A Novel Graph based Trajectory Predictor with Pseudo Oracle

Phase Space Reconstruction Network for Lane Intrusion Action Recognition

Contact Info

Product

Resources

About