Improving Multi-frame Data Association with Sparse Representations for Robust Near-online Multi-object Tracking.
Loïc Fagot-BouquetRomaric AudigierYoann DhomeFrédéric LeraslePublished in: ECCV (8) (2016)
Keyphrases
- multi object tracking
- data association
- sparse representation
- multi frame
- object tracking
- robust tracking
- kalman filter
- computationally efficient
- visual tracking
- motion model
- motion estimation
- particle filter
- object detection
- point correspondences
- computer vision systems
- multi sensor
- image classification
- multi camera
- particle filtering
- image sequences
- multiple objects
- super resolution
- high dimensional data
- dimensionality reduction
- data sets
- optical flow
- video sequences
- three dimensional
- signal processing
- body parts
- feature selection
- reinforcement learning
- computer vision
- face recognition
- multiscale
- mean shift
- multiresolution
- image features