ENBIS-18 in Nancy
2 – 25 September 2018; Ecoles des Mines, Nancy (France)
Abstract submission: 20 December 2017 – 4 June 2018
Self-Similarity Analysis of Electrodermal Activity for Driver’s Stress Level Characterisation
4 September 2018, 11:40 – 12:00
- Submitted by
- Jean-Michel Poggi
- Jean-Michel Poggi (LMO, University of Paris Sud - Orsay and University Paris Descartes), Neska El Haouij (ENIT and University of Tunis El Manar, Tunisia), Raja Ghozi (ENIT and University of Tunis El Manar), Sylvie Sevestre Ghalila (CEA-LinkLab), Mériem Jaïdane (ENIT and University of Tunis El Manar)
- This work characterizes ``stress''' levels via a self-similarity analysis of the Electrodermal Activity (EDA). For that, the Fractional Brownian Motion (FBM), parameterized via the Hurst exponent H, is evoked to model the EDA changes in a real-world driving context.
To characterise the EDA scale invariance, the FBM process and its corresponding exponent H, estimated thanks to a wavelet-based approach, are used. Specifically, an automatic scale range selection is proposed in order to detect the linearity in the logscale diagram. The procedure is applied to the EDA signals, from the open database drivedb, captured originally on the foot and the hand of the drivers during a real-world driving experiment designed to evoke different levels of arousal and stress.
The estimated Hurst exponent H offers a distinction in stress levels when driving in highway versus city, with a reference to restful state of minimal stress level. Specifically, the estimated H decreases when the environmental complexity increases. In addition, almost all the estimated values are greater than 0.5 suggesting that the EDA signal has a long-range dependence. Furthermore, the H estimated on the Foot EDA signals allows a better characterisation of the driving task than the Hand EDA.
This self-similarity analysis captures the complexity the complexity of the EDA signal. Such analysis was applied to various physiological signals in literature but not to the EDA, a signal which was found to correlate most with human affect. The proposed analysis could be useful in real-time monitoring of ``stress'' and arousal levels in urban driving spaces.
Return to programme