Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Prototypical Network with an Attention-based Encoder for Drivers Identification Application

Wei-Hsun Lee, Che-Yu Chang, Kuang-Yu Li

TL;DR

This work tackles driver identification under privacy and data-availability constraints by introducing AttEnc, a lightweight attention-based encoder that achieves state-of-the-art accuracy with far fewer parameters. It further extends this approach with P-AttEnc, a prototypical network that enables few-shot learning to generalize to unseen drivers. The results show AttEnc attaining around 99% accuracy across three driving datasets and significantly reducing model size and inference time, while P-AttEnc demonstrates robust performance in 1- to 10-shot scenarios and the ability to classify unknown drivers. The proposed framework holds practical promise for anti-theft, usage-based insurance, and fleet management with low data requirements and flexible deployment across diverse data sources.

Abstract

Driver identification has become an area of increasing interest in recent years, especially for data- driven applications, because biometric-based technologies may incur privacy issues. This study proposes a deep learning neural network architecture, an attention-based encoder (AttEnc), which uses an attention mechanism for driver identification and uses fewer model parameters than current methods. Most studies do not address the issue of data shortages for driver identification, and most of them are inflexible when encountering unknown drivers. In this study, an architecture that combines a prototypical network and an attention-based encoder (P-AttEnc) is proposed. It applies few-shot learning to overcome the data shortage issues and to enhance model generalizations. The experiments showed that the attention-based encoder can identify drivers with accuracies of 99.3%, 99.0% and 99.9% in three different datasets and has a prediction time that is 44% to 79% faster because it significantly reduces, on average, 87.6% of the model parameters. P-AttEnc identifies drivers based on few shot data, extracts driver fingerprints to address the issue of data shortages, and is able to classify unknown drivers. The first experiment showed that P-AttEnc can identify drivers with an accuracy of 69.8% in the one-shot scenario. The second experiment showed that P-AttEnc, in the 1-shot scenario, can classify unknown drivers with an average accuracy of 65.7%.

A Prototypical Network with an Attention-based Encoder for Drivers Identification Application

TL;DR

This work tackles driver identification under privacy and data-availability constraints by introducing AttEnc, a lightweight attention-based encoder that achieves state-of-the-art accuracy with far fewer parameters. It further extends this approach with P-AttEnc, a prototypical network that enables few-shot learning to generalize to unseen drivers. The results show AttEnc attaining around 99% accuracy across three driving datasets and significantly reducing model size and inference time, while P-AttEnc demonstrates robust performance in 1- to 10-shot scenarios and the ability to classify unknown drivers. The proposed framework holds practical promise for anti-theft, usage-based insurance, and fleet management with low data requirements and flexible deployment across diverse data sources.

Abstract

Driver identification has become an area of increasing interest in recent years, especially for data- driven applications, because biometric-based technologies may incur privacy issues. This study proposes a deep learning neural network architecture, an attention-based encoder (AttEnc), which uses an attention mechanism for driver identification and uses fewer model parameters than current methods. Most studies do not address the issue of data shortages for driver identification, and most of them are inflexible when encountering unknown drivers. In this study, an architecture that combines a prototypical network and an attention-based encoder (P-AttEnc) is proposed. It applies few-shot learning to overcome the data shortage issues and to enhance model generalizations. The experiments showed that the attention-based encoder can identify drivers with accuracies of 99.3%, 99.0% and 99.9% in three different datasets and has a prediction time that is 44% to 79% faster because it significantly reduces, on average, 87.6% of the model parameters. P-AttEnc identifies drivers based on few shot data, extracts driver fingerprints to address the issue of data shortages, and is able to classify unknown drivers. The first experiment showed that P-AttEnc can identify drivers with an accuracy of 69.8% in the one-shot scenario. The second experiment showed that P-AttEnc, in the 1-shot scenario, can classify unknown drivers with an average accuracy of 65.7%.
Paper Structure (15 sections, 7 equations, 8 figures, 2 tables)

This paper contains 15 sections, 7 equations, 8 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Problem Description
  3. Motivation
  4. Related Works
  5. Methodology
  6. Attention-based Encoder (ATTENC)
  7. Prototypical Network with Attention-based Encoder (P-ATTENC)
  8. Experiments
  9. Data Description
  10. Data Preprocessing
  11. Experiment Setup
  12. Evaluations
  13. Result of Experiment (Stage 1)
  14. Result of Experiment (Stage 2)
  15. Conclusion and Future Works

Figures (8)

  • Figure 1: Two-stage classification is used to solve the issues of driver classification and data shortages.
  • Figure 2: Attention-based Encoder for driver identification.
  • Figure 3: Prototypical Network with Attention-based Encoder for Few-Shot Learning.
  • Figure 4: Performance Evaluation in Four Driver Identification Models: Parameters Comparisons
  • Figure 5: Performance Evaluation in Four Driver Identification Models: CPU time
  • ...and 3 more figures