Multimodal Deep Generative Models for Trajectory Prediction: A Conditional Variational Autoencoder Approach

Ivanovic, Boris; Leung, Karen; Szczerbicki, Edward; Pavone, Marco

doi:10.1109/lra.2020.3043163

Cited by 60 publications

(33 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where C(• ∥ •) is the Kullback-Leibler divergence between two distributions and can not be negative. Now, the ELBO can be optimized via stochastic gradient descent algorithm [19], [20]. IV.…”

Section: A Cvae For Parameter Identificationmentioning

confidence: 99%

WECC Composite Load Model Parameter Identification Using Deep Learning Approach

Khazeiynasab¹

2022

Preprint

View full text Add to dashboard Cite

<div>The increased integration of distributed energy resources (DERs) has significantly changed the load composition. As compared to traditional static or induction motor loads, these inverter-based resources (IBRs) exhibit different behavior. Then, conventional dynamic load modeling with ZIP and induction motors (ZIP + IM) is no longer adequate to address the current load characteristic transitions. Western Electricity Coordinating Council (WECC) has developed the composite load model with distributed generation (CMPLDWG) to accurately represent static load, induction motors, and IBRs. In this paper, we propose a novel deep learning-based method using a conditional variational autoencoder (CVAE) to estimate the CMPLDWG model’s posterior distributions. The effectiveness of the proposed method is validated using an IEEE 39-bus test system. The results show that the proposed approach can accurately and efficiently estimate the parameters’ posterior distributions.</div>

show abstract

Section: A Cvae For Parameter Identificationmentioning

confidence: 99%

WECC Composite Load Model Parameter Identification Using Deep Learning Approach

Khazeiynasab¹

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…As the RNN, LSTM 6 and GRU 7 can well extract the time relationship between data features, it has been widely used in many sequence prediction tasks in the field, including speech recognition, 8 machine translation, 9 and trajectory prediction. 10 For the trajectory prediction, the encoder-decoder is used in some methods [11][12][13] to generate the probability distribution information of the vehicle's future position on the occupied raster map, but it can only suitable for a single scene with no other traffic participants and the clearly visible lane lines. Considering the impact of the static environment, Scene-LSTM 14 and SS-LSTM 15 used CNN to extract the bird's-eye view features of the road environment to promote the prediction.…”

Section: Related Workmentioning

confidence: 99%

Probabilistic trajectory prediction of heterogeneous traffic agents based on layered spatio-temporal graph

Zhang

et al. 2021

Proceedings of the Institution of Mechanical Engineers, Part D:

View full text Add to dashboard Cite

In order to safely and comfortably navigate in the complex urban traffic, it is necessary to make multi-modal predictions of autonomous vehicles for the next trajectory of various traffic participants, with the continuous movement trend and inertia of the surrounding traffic agents taken into account. At present, most trajectory prediction methods focus on prediction on future behavior of traffic agents but with limited, consideration of the response of traffic agents to the future behavior of the ego-agent. Moreover, it can only predict the trajectory of single-type agents, which make it impossible to learn interaction in a complex environment between traffic agents. In this paper, we proposed a graph-based heterogeneous traffic agents trajectory prediction model LSTGHP, which consists of the following three parts: (1) layered spatio-temporal graph module; (2) ego-agent motion module; (3) trajectory prediction module, which can realize multi-modal prediction of future trajectories of traffic agents with different semantic categories in the scene. To evaluate its performance, we collected trajectory datasets of heterogeneous traffic agents in a time-varying, highly dynamic urban intersection environment, where vehicles, bicycles, and pedestrians interacted with each other in the scene. It can be drawn from experimental results that our model can improve its prediction accuracy while interacting at a close range. Compared with the previous prediction methods, the model has less prediction error in the trajectory prediction of heterogeneous traffic agents.

show abstract

“…The generative adversarial networks [ 8 ] is another approach for spatiotemporal prediction. A conditional variational autoencoder method has been proposed in [ 9 ] by producing future human trajectories conditioned on previous observations and future robot actions. The prediction methods [ 8 , 9 ] aim to generate less blurry frames, but their performance significantly depends on the unstable training process.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Some methods [ 6 , 7 ] have achieved accurate prediction results, but they cause representation loss. The method of adversarial has been applied in prediction tasks [ 8 , 9 ]. However, they [ 8 , 9 ] are significantly dependent on the unstable training process.…”

Section: Introductionmentioning

confidence: 99%

Dual Memory LSTM with Dual Attention Neural Network for Spatiotemporal Prediction

Guan

2021

Sensors

View full text Add to dashboard Cite

Spatiotemporal prediction is challenging due to extracting representations being inefficient and the lack of rich contextual dependences. A novel approach is proposed for spatiotemporal prediction using a dual memory LSTM with dual attention neural network (DMANet). A new dual memory LSTM (DMLSTM) unit is proposed to extract the representations by leveraging differencing operations between the consecutive images and adopting dual memory transition mechanism. To make full use of historical representations, a dual attention mechanism is designed to capture long-term spatiotemporal dependences by computing the correlations between the current hidden representations and the historical hidden representations from temporal and spatial dimensions, respectively. Then, the dual attention is embedded into DMLSTM unit to construct a DMANet, which enables the model with greater modeling power for short-term dynamics and long-term contextual representations. An apparent resistivity map (AR Map) dataset is proposed in this paper. The B-spline interpolation method is utilized to enhance AR Map dataset and makes apparent resistivity trend curve continuous derivative in the time dimension. The experimental results demonstrate that the developed method has excellent prediction performance by comparisons with some state-of-the-art methods.

show abstract

Multimodal Deep Generative Models for Trajectory Prediction: A Conditional Variational Autoencoder Approach

Cited by 60 publications

References 26 publications

WECC Composite Load Model Parameter Identification Using Deep Learning Approach

WECC Composite Load Model Parameter Identification Using Deep Learning Approach

Probabilistic trajectory prediction of heterogeneous traffic agents based on layered spatio-temporal graph

Dual Memory LSTM with Dual Attention Neural Network for Spatiotemporal Prediction

Contact Info

Product

Resources

About