Psychological distraction is very dangerous in driving. We developed a SDA-based driving distraction approach that discovers prominent features from raw spectrum of EEG data using SDA. This program is a demo how the detection system works.
The package implements two classifiers to discriminate normal and distraction status at the back-end (i.e., when receiving the signal from SDA feature extraction), the first is a one-class SVM, and the second is a two-class SVM.
-
Choose which driver(subject) to detect
-
Choose the type of SVM (one-class SVM or two-class SVM)
-
Train & save model with the driver's EEG data (with the chosen type of SVM)
-
Send the driver's EEG data to Data Analyzer in real-time
-
Receive EEG data of the driver from Data Sender
-
Judge the driver's status using the trained model
-
Display the driver's status and EEG data in real time
- You can get access to the details of the program in the document named 'Document for Demo Program.pdf'.
- More information can be found in the associated paper 'Sparse Discriminative Analysis and Its Application in Distraction Classification'.
Apache 2.0
Yanqing Wang (wangyanqing@cslt.riit.tsinghua.edu.cn) Dong Wang (wangdong99@mails.tsinghua.edu.cn)