# Detecting inter-sectional accuracy differences in driver drowsiness detection algorithms

## TL;DR

This paper evaluates the racial bias in driver drowsiness detection CNNs, highlighting overfitting issues and proposing a PCA-based visualization method to identify potential discrimination in model performance across different ethnicities.

## Contribution

It demonstrates the racial bias present in existing datasets and models, and introduces a novel PCA visualization technique to detect potential discrimination.

## Key findings

- Models trained on public datasets show overfitting and racial bias.
- Testing on diverse datasets reveals significant accuracy disparities.
- The PCA visualization helps identify groups at risk of discrimination.

## Abstract

Convolutional Neural Networks (CNNs) have been used successfully across a broad range of areas including data mining, object detection, and in business. The dominance of CNNs follows a breakthrough by Alex Krizhevsky which showed improvements by dramatically reducing the error rate obtained in a general image classification task from 26.2% to 15.4%. In road safety, CNNs have been applied widely to the detection of traffic signs, obstacle detection, and lane departure checking. In addition, CNNs have been used in data mining systems that monitor driving patterns and recommend rest breaks when appropriate. This paper presents a driver drowsiness detection system and shows that there are potential social challenges regarding the application of these techniques, by highlighting problems in detecting dark-skinned driver's faces. This is a particularly important challenge in African contexts, where there are more dark-skinned drivers. Unfortunately, publicly available datasets are often captured in different cultural contexts, and therefore do not cover all ethnicities, which can lead to false detections or racially biased models. This work evaluates the performance obtained when training convolutional neural network models on commonly used driver drowsiness detection datasets and testing on datasets specifically chosen for broader representation. Results show that models trained using publicly available datasets suffer extensively from over-fitting, and can exhibit racial bias, as shown by testing on a more representative dataset. We propose a novel visualisation technique that can assist in identifying groups of people where there might be the potential of discrimination, using Principal Component Analysis (PCA) to produce a grid of faces sorted by similarity, and combining these with a model accuracy overlay.