Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Master Thesis: Drivers’ affective state detection and regulation
You will explore the possibility for personalized in-car services of the future. Recently, we collected a unique dataset, consisting of recorded driving behavior in the wild. In the term of your master thesis, you will focus on finding patterns within this dataset to identify driver states.
Keywords: machine learning, car driving, data analysis, gps, audio processing, video processing, can data
The improvement of vehicular technologies has equipped cars with an intelligence that makes many new applications possible. The car as a modern commodity for many people provides a controlled environment in which they spend a substantial amount of time. We explore whether AI in vehicles can be leveraged for analyzing people’s health status by monitoring and evaluating their driving behaviors.
During our most recent field study, we collected a unique dataset which will be the foundation for your thesis.
The dataset consists of:
- The data was collected over 4 months by people driving in the wild
- 50+ CAN/car sensors
- Video (inside the car and the street in front)
- Audio recordings of drivers
- HRV values from semiprofessional and wearable devices
- A vast amount of completed surveys about a driver’s state (with focus on affect and emotion)
- Interventions conducted with the drivers while driving
More information about the project and our research goals:
- https://www.iot-lab.ch/projects-connectedmobility/arne/
- Unfortunately only in German: https://www.iot-lab.ch/news/aufruf-probanden-suche-fur-feldstudie-entspannung-durch-autofahren/
The improvement of vehicular technologies has equipped cars with an intelligence that makes many new applications possible. The car as a modern commodity for many people provides a controlled environment in which they spend a substantial amount of time. We explore whether AI in vehicles can be leveraged for analyzing people’s health status by monitoring and evaluating their driving behaviors.
During our most recent field study, we collected a unique dataset which will be the foundation for your thesis.
The dataset consists of:
- The data was collected over 4 months by people driving in the wild
- 50+ CAN/car sensors
- Video (inside the car and the street in front)
- Audio recordings of drivers
- HRV values from semiprofessional and wearable devices
- A vast amount of completed surveys about a driver’s state (with focus on affect and emotion)
- Interventions conducted with the drivers while driving
More information about the project and our research goals:
- Unfortunately only in German: https://www.iot-lab.ch/news/aufruf-probanden-suche-fur-feldstudie-entspannung-durch-autofahren/
We are interested in using our available dataset to predict driver states. Accordingly, we target to establish several data analysis streams which lead to the following research questions:
How can we predict a driver’s state by the means of…
- … facial expressions based on a driver camera?
- … CAN (car data)?
- … inside audio recordings?
- … adding additional annoted information to the data? (e.g. Open Street Map information, creating a trip log book,...)
- … your idea based on the described data above.
Your task will be to design a complete preprocessing, analysis, and prediction machine learning pipeline. During the master thesis, you will work closely with us and will receive a dedicated supervision. We are highly interested to invest time in your thesis as the topics above count into our research.
We are interested in using our available dataset to predict driver states. Accordingly, we target to establish several data analysis streams which lead to the following research questions:
How can we predict a driver’s state by the means of…
- … facial expressions based on a driver camera?
- … CAN (car data)?
- … inside audio recordings?
- … adding additional annoted information to the data? (e.g. Open Street Map information, creating a trip log book,...)
- … your idea based on the described data above.
Your task will be to design a complete preprocessing, analysis, and prediction machine learning pipeline. During the master thesis, you will work closely with us and will receive a dedicated supervision. We are highly interested to invest time in your thesis as the topics above count into our research.
Kevin Koch (kevinkoch@ethz.ch) or Shu Liu (liush@ethz.ch).
Please contact us with your CV, a short statement of motivation, and your current transcripts of records (bachelor & master).
Kevin Koch (kevinkoch@ethz.ch) or Shu Liu (liush@ethz.ch).
Please contact us with your CV, a short statement of motivation, and your current transcripts of records (bachelor & master).