Context-Aware Driver Emotion Analysis for Intelligent Transportation Systems
Year: 2026
Type: article
Abstract: In the transportation system, road safety is a primary area of concern. A multitude of factors contribute towards the road safety wherein driver’s emotion plays a significant role. By inspecting the driver’s emotional behaviour, Intelligent Transportation System (ITS) through its monitoring systems, majorly aims to maintain safety vigilance. However, the driver’s emotions are predominantly assessed using various sensors which can be intrusive in nature and can make the driver self-conscious leading to a skewed prediction. To avoid this, we have considered different contextual factors (weather, road condition, acceleration, deceleration etc.) in this proposed work that contributes towards the driver’s emotions getting affected which can be due to both external and internal conditions on the road. The information regarding the contextual factors have been gathered through questionnaire responses under three major categories namely the vehicle, surrounding environment, and driver itself. Considering these contexts in totality, a dataset is curated onto which we have trained a proposed 1D-CNN model. To improve the model’s performance, various resampling techniques have been implemented wherein Synthetic Minority Over-sampling Technique SMOTE + ENN yields best accuracy of 99.78%. To substantiate the proposed model’s predictions, a 2D-CNN model has been implemented for verifying the driver’s emotion through the facial images. Furthermore, the proposed model has been validated using a real-time sample questionnaire response and the corresponding driver’s facial image for predicting the driver’s emotion on drive. The novelty of the proposed approach lies in the driver emotion identification considering the complete context and auxiliary validation through facial images.
Authors: Anwesha Patel, Rishu Chhabra, C. Rama Krishna
Language: English
Topic: Emotion and Mood Recognition
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Keywords: driver emotion recognition, intelligent transportation systems, road safety, contextual factors, weather conditions, road conditions, vehicle acceleration and deceleration, 1D convolutional neural network, SMOTE, Enns, class imbalance, facial emotion recognition, 2D convolutional neural network, questionnaire data
Open Access status: hybrid
Estimated APC paid: $3,690
OpenAlex ID: W7235042314
