Direction-Decision Learning Based Pedestrian Flow Behavior Investigation

Zhang, Zhe; Jia, Limin

doi:10.1109/access.2020.2964001

Cited by 11 publications

(9 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Physics and Neural Network mixed approaches. Past attempts to use machine learning techniques in combination with physical models have applied gradient descend methods to learn the interaction forces [21], used linear regression and NN to predict the direction of motion [22], used evolutionary algorithm to optimise the SFM parameters from video segments [23]. The approach taken in this paper is rather new and is inspired to the simulation of vehicle dynamics by [24].…”

Section: A Related Workmentioning

confidence: 99%

Efficient Prediction of Human Motion for Real-Time Robotics Applications With Physics-Inspired Neural Networks

et al. 2022

View full text Add to dashboard Cite

Generating accurate and efficient predictions for the motion of the humans present in the scene is key to the development of effective motion planning algorithms for robots moving in promiscuous areas, where wrong planning decisions could generate safety hazard or simply make the presence of the robot "socially" unacceptable. Our approach to predict human motion is based on a neural network of a peculiar kind. Contrary to conventional deep neural networks, our network embeds in its structure the popular Social Force Model, a dynamic equation describing the motion in physical terms. This choice allows us to concentrate the learning phase in the aspects which are really unknown (i.e., the model's parameters) and to keep the structure of the network simple and manageable. As a result, we are able to obtain a good prediction accuracy even by using a small and synthetically generated training set. Importantly, the prediction accuracy remains acceptable even when the network is applied in scenarios radically different from those for which it was trained. Finally, the choices of the network are "explainable", as they can be interpreted in physical terms. Comparative and experimental results prove the effectiveness of the proposed approach.

show abstract

Section: A Related Workmentioning

confidence: 99%

Efficient Prediction of Human Motion for Real-Time Robotics Applications With Physics-Inspired Neural Networks

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Two major factors affect the route choice: (a) environmental and (b) personal factors. e most influential environmental factors are distances from destination [24], congestion degree [25], speed difference [26], and exit's width [27,28]. Apart from environmental factors, personal factors can also affect exit choice.…”

Section: Literature Reviewmentioning

confidence: 99%

Modeling and Simulation of Departure Passenger’s Behavior Based on an Improved Social Force Approach: A Case Study on an Airport Terminal in China

Zhang

Kong

et al. 2021

Advances in Civil Engineering

View full text Add to dashboard Cite

The unprecedented growth of passenger throughput in large airport terminals highlights the importance of analyzing passengers’ movement to achieve airport terminal’s elaborate management. Based on the theory of original social force model, video data from a departure hall of a large airport terminal in China were analyzed to summarize passengers’ path planning characteristics. Then, a double-level model was established to describe passengers’ path planning behaviors. At the decision level of the proposed model, the avoiding force model including common avoiding force and additional horizontal avoiding force was established on the basis of setting time and space limitations for taking avoiding action and was used to describe passengers’ path planning in close-range space. At the tactical level of the proposed model, the route and node choice models were established to describe passengers’ path planning in long-range space. In the route choice model, a distribution of intermediate destination areas was proposed, with detouring distance, pedestrian density, speed difference, and pedestrian distribution considered in choosing an intermediate destination area. In the node choice model, the walking distance, the quantity of people waiting, and luggage were considered in choosing a check-in counter or security check channel. The main parameters of the proposed model were confirmed according to video data. Simulation results show that the proposed model can simulate departure passengers’ path planning behaviors at an acceptable accuracy level.

show abstract

“…Such a mechanism allows notably to explain clogging effects at bottlenecks. Recent decision-based models are based on cognitive effect [124], [125] or learning process [37].…”

Section: B Microscopic Pedestrian Modelsmentioning

confidence: 99%

“…Examples of the hybrid approach are given in applications like the discovering of novel climate patterns [248], [249], the finding of novel compounds in material science [250], [251], the designing density functionals in quantum chemistry [252], or the improving imaging technologies in bio-medical science [253], [254]. Recently, the first applications for pedestrian trajectory predictions can be found in [255], [34], [36].…”

Section: A the Hybrid Approachmentioning

confidence: 99%

“…Furthermore, another possibility is to use DL methods to improve the accuracy of the KB models. This can be done by estimating the parameters based on the results of an DL algorithm or to implement a neural network in the KB simulations [37], [38]. These combinations are called the hybrid approach and they benefit from the strengths of both approaches and avoid their shortcomings.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Review of Pedestrian Trajectory Prediction Methods: Comparing Deep Learning and Knowledge-based Approaches

Korbmacher¹,

Tordeux²

2021

Preprint

View full text Add to dashboard Cite

In crowd scenarios, predicting trajectories of pedestrians is a complex and challenging task depending on many external factors. The topology of the scene and the interactions between the pedestrians are just some of them. Due to advancements in data-science and data collection technologies deep learning methods have recently become a research hotspot in numerous domains. Therefore, it is not surprising that more and more researchers apply these methods to predict trajectories of pedestrians. This paper compares these relatively new deep learning algorithms with classical knowledge-based models that are widely used to simulate pedestrian dynamics. It provides a comprehensive literature review of both approaches, explores technical and application oriented differences, and addresses open questions as well as future development directions. Our investigations point out that the pertinence of knowledge-based models to predict local trajectories is nowadays questionable because of the high accuracy of the deep learning algorithms. Nevertheless, the ability of deep-learning algorithms for large-scale simulation and the description of collective dynamics remains to be demonstrated. Furthermore, the comparison shows that the combination of both approaches (the hybrid approach) seems to be promising to overcome disadvantages like the missing explainability of the deep learning approach.

show abstract

Direction-Decision Learning Based Pedestrian Flow Behavior Investigation

Cited by 11 publications

References 41 publications

Efficient Prediction of Human Motion for Real-Time Robotics Applications With Physics-Inspired Neural Networks

Efficient Prediction of Human Motion for Real-Time Robotics Applications With Physics-Inspired Neural Networks

Modeling and Simulation of Departure Passenger’s Behavior Based on an Improved Social Force Approach: A Case Study on an Airport Terminal in China

Review of Pedestrian Trajectory Prediction Methods: Comparing Deep Learning and Knowledge-based Approaches

Contact Info

Product

Resources

About