uMoDT: an unobtrusive multi-occupant detection and tracking using robust Kalman filter for real-time activity recognition.
Muhammad Asif RazzaqJavier Medina QueroIan ClelandChris D. NugentUsman AkhtarHafiz Syed Muhammad BilalUbaid Ur RehmanSungyoung LeePublished in: Multim. Syst. (2020)
Keyphrases
- machine learning
- kalman filter
- activity recognition
- robust tracking
- real time
- smart home
- kalman filtering
- object tracking
- activity analysis
- tracking framework
- partial occlusion
- particle filter
- human activity recognition
- visual surveillance
- human activities
- mean shift
- action recognition
- target tracking
- particle filtering
- state estimation
- visual tracking
- accelerometer data
- computer vision
- ambient intelligence
- extended kalman filter
- moving target
- data association
- appearance model
- anomaly detection
- state space model
- image sequences
- video surveillance
- image classification
- data sets