Detecting Stressful Social Interactions Using Wearable Physiological and Inertial Sensors
summary
Lab and field data from 38 cohabiting participants wearing a chest patch, wristband and phone are used to label and classify stressors; the team defines stress cycles (rise, saturation, decay, recovery) and extracts morphological features plus wrist motion energy that is higher during social stress. An algorithm locates cycle points with 96% point detection, a random forest using cycle features plus motion reaches F1 = 0.83 (one-cycle detection yields F1 ≈ 0.74), enabling intervention within about 3.9 minutes.
AI-written summary · about
related talks
06mPredicting Subjective Measures of Social Anxiety from Sparsely Collected Mobile Sensor Data
17mEMPress: Practical Hand Gesture Classification with Wrist-Mounted EMG and Pressure Sensing
09mAn Activity Recognition System for Taking Medicine Using In-The-Wild Data to Promote Medication ...
01h 13mHealth Monitoring with Wireless Sensors and Machine Learning
27mHARSHA URLAM & SWETA SINGH at Cypher 2017
06mQuantifying the Relationships between Everyday Objects and Emotional States through Deep ...
22mSeeing, Sensing and Recognizing Laban Movement Qualities
19mInTheMoodForLife: Analyzing Sleep to Better Manage Mood Disorders