The Potential of Cognitive Circles to Measure Mental Load
summary
Finger kinematics from tracing circles on a tablet while doing arithmetic, reading, or spatial tasks predict task difficulty: 40 participants yield 240 two‑minute series for training and 8 participants 48 series for testing. Features include X/Y, velocity, acceleration and angular velocity; models tested: LSTM on raw series, Random Forest on channel means and MiniRocket, and MobileNet on trace images. Results show 75% average accuracy, up to 94% for spatial reasoning and near 50% for reading, so per‑task calibration is required.
AI-written summary · about
chapters (7)
- 0:00Introduction to cognitive circles
- 1:48Traditional workload measurement methods
- 3:13Experiment design and task types
- 6:49Data representations and models
- 7:51Results and accuracy findings
- 9:14Implications for HCI and future work
- 10:40Q&A: Model transferability
related talks
06mObjective Measures of Cognitive Load Using Deep Multi-Modal Learning — A Use-Case in Aviation
07mAssessing Cognitive Performance Using Physiological and Facial Features: Generalizing Across ...
26mBrainDrain Using Machine Learning and Brain Waves to Detect Errors in Human
17mLearning to assess the quality of stroke rehabilitation exercises
07mMultimodal Emotion Recognition of Hand-Object Interaction
22mSeeing, Sensing and Recognizing Laban Movement Qualities
03mIdentifying Children Metacognitive Monitoring Performance Through Facial Expressions
08mSo Predictable! Continuous 3D Hand Trajectory Prediction in Virtual Reality