VEmotion: Using Driving Context for Indirect Emotion Prediction in Real-Time
summary
A smartphone-only system predicts driver emotions from contextual signals (GPS speed and acceleration, road type, speed limit, real-time traffic, weather, daytime, age and prior felt emotion) and aligns streams on-device for real-time classification. Evaluated in the wild with N=12 participants (mean route 16 minutes), the model achieves 71.7% accuracy, outperforming facial-expression baselines by about 29% person-dependent (and ~8.5% person-independent). A real-time app runs locally, and adding face data does not improve results.
AI-written summary · about
related talks
06mThe Empathetic Car: Exploring Emotion Inference via Driver Behaviour and Traffic Context
11mAmbient Intelligence and Emotionally Intelligent Devices
34mEnhancing Customer Experiences with Emotional AI,
49mAffective Computing – Bringing Humans and Machines Closer Through Emotions
15mRecognizing Emotional Expression in Game Streams
50mAffective Computing
19mWhat Can Be Predicted from Six Seconds of Driver Glances?
21mEmotion Estimation for Social Robotics