Reliable object tracking by multimodal hybrid feature extraction and transformer-based fusion.
Hongze SunRui LiuWuque CaiJun WangYue WangHuajin TangYan CuiDezhong YaoDaqing GuoPublished in: Neural Networks (2024)
Keyphrases
- object tracking
- feature extraction
- particle filter
- mean shift
- video sequences
- image sequences
- kalman filter
- moving objects
- appearance model
- video surveillance
- occlusion handling
- data fusion
- multiple cameras
- partial occlusion
- particle filtering
- tracking objects
- feature selection
- visual tracking
- visual object tracking
- multiple object tracking
- neural network
- sparse representation
- image classification
- robust object tracking
- video surveillance systems
- computer vision
- face recognition
- scene segmentation
- principal component analysis
- appearance modeling
- multi sensor
- image processing
- multi camera
- robust tracking
- reinforcement learning
- pattern recognition
- dimensionality reduction
- object location
- tracking accuracy
- feature space
- feature vectors
- machine learning