Driving Style Profiling using Deep Auto-Encoders for Safety Applications in Urban and Highway Scenarios

This paper discusses driving style profiling using deep auto-encoders to enhance safety in urban and highway scenarios. The proposed method leverages sensor data from vehicles to create unique driving profiles, using advanced techniques such as Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNNs). By analyzing temporal dependencies and deviations from standard driving behaviors, the approach aims to improve vehicular safety across diverse scenarios.

Recommended citation: Dhanya Krishnan, Divya Srinivasan, Balasubramanian V, Nagaradjane Prabagarane, Joannes Sam Mertens, Salvatore D. Cafiso, Laura Galluccio, Giacomo Morabito, Giuseppina Pappalardo. (2024). "Driving Style Profiling using Deep Auto-Encoders for Safety Applications in Urban and Highway Scenarios." IEEE International Symposium on Measurements & Networking (M&N).
Download Paper | Download Slides