TAO: A Large-Scale Benchmark for Tracking Any Object.
Achal DaveTarasha KhuranaPavel TokmakovCordelia SchmidDeva RamananPublished in: ECCV (5) (2020)
Keyphrases
- object tracking
- deformable objects
- partial occlusion
- target object
- object position
- real world
- multiple objects
- d objects
- small scale
- complex objects
- tracking objects
- particle filter
- articulated objects
- real time
- visual tracking
- object detection and tracking
- rigid objects
- object matching
- data objects
- tracking framework
- moving objects
- particle filtering
- appearance variations
- camera views
- cluttered background
- computer vision
- mean shift
- motion tracking
- appearance model
- kalman filter
- moving target
- bounding box
- image sequences
- object recognition
- partially occluded
- object segmentation
- video surveillance
- point correspondences
- object location
- motion model