Real Time Eye Tracking Data Acquisition And Preprocessing

via “real-time facial landmark detection and tracking”

LivePortrait — AI demo on HuggingFace

Unique: Implements temporal smoothing through a learned motion model rather than post-hoc filtering, reducing jitter while preserving fast expression changes by predicting landmark positions based on optical flow and previous frame history

vs others: Achieves lower latency than MediaPipe for video processing and higher accuracy than traditional Dlib-based methods because it uses modern transformer architectures with temporal context aggregation

via “real-time-eye-tracking-data-acquisition-and-preprocessing”

Unique: Implements hardware-specific calibration and real-time preprocessing pipelines (blink detection, saccade detection, fixation clustering) optimized for clinical eye-tracking hardware, with quality assurance metrics that validate tracking fidelity before data enters clinical analysis workflows

vs others: Provides clinical-grade eye-tracking data acquisition with real-time quality validation, superior to consumer-grade eye-tracking (e.g., webcam-based gaze estimation) which lacks hardware calibration, has 2-5x lower accuracy, and cannot reliably detect saccades or fixations