Model-predictive planning for autonomous vehicles anticipating intentions of vulnerable road users by artificial neural networks

Eilbrecht, Jan; Bieshaar, Maarten; Zernetsch, Stefan; Doll, Konrad; Sick, Bernhard; Stursberg, Olaf

doi:10.1109/ssci.2017.8285249

Cited by 15 publications

(9 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, we plan to validate our algorithm by incorporating probabilistic forecasts into an existing planning algorithm [15].…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Cyclist Intention Detection: A Probabilistic Approach

Zernetsch¹,

Reichert²,

Kress³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

This article presents a holistic approach for probabilistic cyclist intention detection. A basic movement detection based on motion history images (MHI) and a residual convolutional neural network (ResNet) are used to estimate probabilities for the current cyclist motion state. These probabilities are used as weights in a probabilistic ensemble trajectory forecast. The ensemble consists of specialized models, which produce individual forecasts in the form of Gaussian distributions under the assumption of a certain motion state of the cyclist (e.g. cyclist is starting or turning left). By weighting the specialized models, we create forecasts in the from of Gaussian mixtures that define regions within which the cyclists will reside with a certain probability. To evaluate our method, we rate the reliability, sharpness, and positional accuracy of our forecasted distributions. We compare our method to a single model approach which produces forecasts in the form of Gaussian distributions and show that our method is able to produce more reliable and sharper outputs while retaining comparable positional accuracy. Both methods are evaluated using a dataset created at a public traffic intersection. Our code and the dataset are made publicly available.

show abstract

“…Furthermore, we plan to validate our algorithm by incorporating probabilistic forecasts into an existing planning algorithm [15].…”

Section: Discussionmentioning

confidence: 99%

“…This is often achieved through prediction of regions instead of positions in the form of probability distributions. In [15], the method from [10] was enhanced by an uncertainty estimate created by an unconditional model. The uncertainty estimates were used with a planning method based on predictive control.…”

Section: B Related Workmentioning

confidence: 99%

Cyclist Intention Detection: A Probabilistic Approach

Zernetsch¹,

Reichert²,

Kress³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Gaussian distributions describing the cyclists' future positions were forecasted in [6] using recurrent neural networks. In [7], forecasted trajectories were extended by an uncertainty estimation through an unconditional, constant model and used for path planning of autonomous vehicles. The authors of [8] proposed a method for forecasting all possible future positions of pedestrians in a set-based fashion using reachability analysis, contextual information, and traffic rules.…”

Section: B Related Workmentioning

confidence: 99%

Pose and Semantic Map Based Probabilistic Forecast of Vulnerable Road Users' Trajectories

Kress¹,

Jeske²,

Zernetsch³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

In this article, an approach for probabilistic trajectory forecasting of vulnerable road users (VRUs) is presented, which considers past movements and the surrounding scene. Past movements are represented by 3D poses reflecting the posture and movements of individual body parts. The surrounding scene is modeled in the form of semantic maps showing, e.g., the course of streets, sidewalks, and the occurrence of obstacles. The forecasts are generated in grids discretizing the space and in the form of arbitrary discrete probability distributions. The distributions are evaluated in terms of their reliability, sharpness, and positional accuracy. We compare our method with an approach that provides forecasts in the form of Gaussian distributions and discuss the respective advantages and disadvantages. Thereby, we investigate the impact of using poses and semantic maps. With a technique called spatial label smoothing, our approach achieves reliable forecasts. Overall, the poses have a positive impact on the forecasts. The semantic maps offer the opportunity to adapt the probability distributions to the individual situation, although at the considered forecasted time horizon of 2.52 s they play a minor role compared to the past movements of the VRU. Our method is evaluated on a dataset recorded in inner-city traffic using a research vehicle. The dataset is made publicly available.

show abstract

“…Using the sets computed by QS, we are able to derive statistical guarantees based on historical data, e.g., an area (i.e., QS) associated with 99% probability of residence. These statistically derived bounds are an essential ingredient to make optimal decision under uncertainty, i.e., they are of great interest for safe trajectory planning [38]. In this section, we apply our QS methodology to forecast the future short-term trajectories of cyclists.…”

Section: Cyclist Trajectory Forecastingmentioning

confidence: 99%

“…We model the forecasting problem as a time series auto-regressive task, i.e., we consider features based on the past head trajectory, and we aim to predict the future course of the trajectory. The general forecasting methodology, including the features used and the choice of a location-independent ego-coordinate frame for prediction, is inspired by [38], [39], [40].…”

Section: Cyclist Trajectory Forecastingmentioning

confidence: 99%

Quantile Surfaces -- Generalizing Quantile Regression to Multivariate Targets

Bieshaar,

Schreiber,

Vogt

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

In this article, we present a novel approach to multivariate probabilistic forecasting. Our approach is based on an extension of single-output quantile regression (QR) to multivariate-targets, called quantile surfaces (QS). QS uses a simple yet compelling idea of indexing observations of a probabilistic forecast through direction and vector length to estimate a central tendency. We extend the single-output QR technique to multivariate probabilistic targets. QS efficiently models dependencies in multivariate target variables and represents probability distributions through discrete quantile levels. Therefore, we present a novel two-stage process. In the first stage, we perform a deterministic point forecast (i.e., central tendency estimation). Subsequently, we model the prediction uncertainty using QS involving neural networks called quantile surface regression neural networks (QSNN). Additionally, we introduce new methods for efficient and straightforward evaluation of the reliability and sharpness of the issued probabilistic QS predictions. We complement this by the directional extension of the Continuous Ranked Probability Score (CRPS) score. Finally, we evaluate our novel approach on synthetic data and two currently researched realworld challenges in two different domains: First, probabilistic forecasting for renewable energy power generation, second, shortterm cyclists trajectory forecasting for autonomously driving vehicles. Especially for the latter, our empirical results show that even a simple one-layer QSNN outperforms traditional parametric multivariate forecasting techniques, thus improving the state-of-the-art performance.

show abstract

Model-predictive planning for autonomous vehicles anticipating intentions of vulnerable road users by artificial neural networks

Cited by 15 publications

References 20 publications

Cyclist Intention Detection: A Probabilistic Approach

Cyclist Intention Detection: A Probabilistic Approach

Pose and Semantic Map Based Probabilistic Forecast of Vulnerable Road Users' Trajectories

Quantile Surfaces -- Generalizing Quantile Regression to Multivariate Targets

Contact Info

Product

Resources

About