Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Deep Neural Network Approach to Fare Evasion

Johannes van der Vyver

TL;DR

This work addresses fare evasion in public transport by predicting passenger actions from time-series keypoint data extracted from video, using an LSTM-based CAR framework and proposing ReID integration for enhanced accuracy. The method processes 3-second keypoint windows to classify payments versus evasions, with evaluation via confusion matrices and real-time footage validation. Key findings show the LSTM achieves strong predictive performance and outperforms a baseline MLP, suggesting that a CNN–LSTM–ReID combination could further reduce fare loss and reliance on inspectors. The study has practical implications for deploying automated, real-time fare-duty monitoring in public transit, while noting limitations such as dataset scope and potential keypoint-detection errors, and outlining directions for broader evasion modalities and hardware deployment.

Abstract

Fare evasion is a problem for public transport companies, with LSTM models this issue can help companies get an analytical insight into where this issue occurs the most, to prevent capital loss. In addition to the financial burden this problem causes, having more inspectors is not enough to alleviate the problem. The purpose of this study is to find a different way to predict fare evasion in the public transport sector. Through the use of keypoint extractions of passengers in video footage, an LSTM model is trained on those keypoints to help predict the actions of passengers between payments and evasions. The results were promising when it came to predicting the actions of passengers on real-time footage. Thus a sophisticated approach can help to decrease the fare evasion problem. A ReID model can be used alongside the LSTM model for better accuracy, as there is always the chance that a person might only pay for the fare at a later stage. With both models, it is possible for public transport companies to start narrowing down where the root of their fare evasion problems emerges.

A Deep Neural Network Approach to Fare Evasion

TL;DR

This work addresses fare evasion in public transport by predicting passenger actions from time-series keypoint data extracted from video, using an LSTM-based CAR framework and proposing ReID integration for enhanced accuracy. The method processes 3-second keypoint windows to classify payments versus evasions, with evaluation via confusion matrices and real-time footage validation. Key findings show the LSTM achieves strong predictive performance and outperforms a baseline MLP, suggesting that a CNN–LSTM–ReID combination could further reduce fare loss and reliance on inspectors. The study has practical implications for deploying automated, real-time fare-duty monitoring in public transit, while noting limitations such as dataset scope and potential keypoint-detection errors, and outlining directions for broader evasion modalities and hardware deployment.

Abstract

Fare evasion is a problem for public transport companies, with LSTM models this issue can help companies get an analytical insight into where this issue occurs the most, to prevent capital loss. In addition to the financial burden this problem causes, having more inspectors is not enough to alleviate the problem. The purpose of this study is to find a different way to predict fare evasion in the public transport sector. Through the use of keypoint extractions of passengers in video footage, an LSTM model is trained on those keypoints to help predict the actions of passengers between payments and evasions. The results were promising when it came to predicting the actions of passengers on real-time footage. Thus a sophisticated approach can help to decrease the fare evasion problem. A ReID model can be used alongside the LSTM model for better accuracy, as there is always the chance that a person might only pay for the fare at a later stage. With both models, it is possible for public transport companies to start narrowing down where the root of their fare evasion problems emerges.
Paper Structure (9 sections, 5 figures)

This paper contains 9 sections, 5 figures.

Table of Contents

  1. Introduction
  2. Experimental Design
  3. Datasets
  4. Experimental Setup
  5. Analysis
  6. Tools and Libraries
  7. Experimental Results
  8. Discussion
  9. Conclusion

Figures (5)

  • Figure 1: Categorical Accuracy of the LSTM model
  • Figure 2: Confusion Matrix for Payments of the LSTM model
  • Figure 3: Confusion Matrix for Payments of the Multilayer Perceptron model
  • Figure 4: Payment Predicted with LSTM Model
  • Figure 5: Evasion Predicted with LSTM Model