• Corpus ID: 239998705

Traffic Forecasting on Traffic Moving Snippets

@article{Wiedemann2021TrafficFO,
  title={Traffic Forecasting on Traffic Moving Snippets},
  author={Nina Wiedemann and Martin Raubal},
  journal={ArXiv},
  year={2021},
  volume={abs/2110.14383}
}
Advances in traffic forecasting technology can greatly impact urban mobility. In the traffic4cast competition, the task of short-term traffic prediction is tackled in unprecedented detail, with traffic volume and speed information available at 5 minute intervals and high spatial resolution. To improve generalization to unknown cities, as required in the 2021 extended challenge, we propose to predict small quadratic city sections, rather than processing a full-city-raster at once. At test time… 

Figures and Tables from this paper

References

SHOWING 1-10 OF 22 REFERENCES
Traffic4cast-Traffic Map Movie Forecasting - Team MIE-Lab
TLDR
This document summarizes the efforts that led to the best submission of the IARAI competition traffic4cast, and gives some insights about which other approaches the authors evaluated, and why they did not work as well as imagined.
Traffic4cast at NeurIPS 2020 ? yet more on theunreasonable effectiveness of gridded geo-spatial processes
TLDR
The IARAI Traffic4cast challenge at NeurIPS 2020 sought to both challenge some assumptions inherent in the 2019 competition design and explore how far this neural network technique can be pushed, and found that the prediction horizon can be extended successfully to 60 minutes into the future.
Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting
TLDR
This paper proposes a novel deep learning framework, Spatio-Temporal Graph Convolutional Networks (STGCN), to tackle the time series prediction problem in traffic domain, and builds the model with complete convolutional structures, which enable much faster training speed with fewer parameters.
Graph-ResNets for short-term traffic forecasts in almost unknown cities
TLDR
A ResNet-inspired graph convolutional neural network approach that uses street network-based subgraphs of the image lattice graphs as a prior that formulating the prediction problem on a graph allows the neural network to learn properties given by the underlying street network.
The surprising efficiency of framing geo-spatial time series forecasting as a video prediction task - Insights from the IARAI Traffic4cast Competition at NeurIPS 2019
TLDR
The success and insights obtained in a first Traffic Map Movie forecasting challenge are reported on and key findings are summarized to discuss promising future directions of the Traffic4cast competition at NeurIPS.
Traffic4cast 2020 - Graph Ensemble Net and the Importance of Feature And Loss Function Design for Traffic Prediction
TLDR
The team tackled the challenge to Traffic4cast 2020 by studying the importance of feature and loss function design, and achieved significant improvement to the best performing U-Net solution from last year, and introduced a novel ensemble GNN architecture which outperformed the GNN solution from this year.
Utilizing UNet for the future traffic map prediction task Traffic4cast challenge 2020
TLDR
This paper describes the UNet based experiments on the Traffic4cast challenge 2020, to predict traffic flow volume, direction and speed on a high resolution map of three large cities worldwide.
Applications of deep learning in congestion detection, prediction and alleviation: A survey
  • Nishant Kumar, M. Raubal
  • Computer Science, Mathematics
    Transportation Research Part C: Emerging Technologies
  • 2021
TLDR
This survey presents the current state of deep learning applications in the tasks related to detection, prediction, and alleviation of congestion, and presents some suggestions for future research directions as answers to the identified challenges.
Traffic map prediction using UNet based deep convolutional neural network
TLDR
This paper used UNet based deep convolutional neural network to train predictive model for the short term traffic forecast on the Traffic4cast challenge 2019, and achieved best performance in this challenge.
EVALUATION OF UNet AND UNet++ ARCHITECTURES IN HIGH RESOLUTION IMAGE CHANGE DETECTION APPLICATIONS
TLDR
This work trains and evaluates two CNN architectures, UNet and UNet++, on a change detection task using Very High-Resolution satellite images collected at two different time epochs and examines and analyse the effect of two different loss functions, a combination of the Binary Cross Entropy Loss with the Dice Loss and the Lovasz Hinge loss.
...
1
2
3
...