1 / 5

Using Electroencephalography (EEG) for User State / Task Classification in HCI Research

Using Electroencephalography (EEG) for User State / Task Classification in HCI Research. Desney Tan Microsoft Research In collaboration with: Johnny Lee (Carnegie Mellon U.) Greg Smith, Ed Cutrell, Mary Czerwinski, Eric Horvitz (Microsoft Research). Approach.

aman
Télécharger la présentation

Using Electroencephalography (EEG) for User State / Task Classification in HCI Research

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Using Electroencephalography (EEG) for User State / Task Classification in HCI Research Desney Tan Microsoft Research In collaboration with: Johnny Lee (Carnegie Mellon U.) Greg Smith, Ed Cutrell, Mary Czerwinski, Eric Horvitz (Microsoft Research)

  2. Approach • Use low-cost EEG to classify user state or task • If we pick appropriate states/tasks, we can… • Control computers with thought alone • Evaluate systems and interfaces • Build intelligent adaptive systems Measure EEG signal (labeled with states of interest) Generate and select relevant features Build Model (Bayes Net) Classify new (unlabeled) Data

  3. Random classifier (or human) Experiment 1 • Cognitive tasks in controlled environment • Rest v. Mental math v. Mental object rotation • 84% accuracy!

  4. Experiment 2 • Halo task in ‘real-world’ environment • Rest v. Play alone v. Play against enemy • 92% accuracy

  5. Contributions and Future Work • Low-cost system can be used • Works in ‘real world’ computing environment • Rather than preventing/filtering, let machine treat ‘noise’ as ‘signal’ where appropriate • Characterize states/tasks we can measure • Figure out how to generalize models • Apply to evaluation or adaptive systems

More Related