Multi-Scale Recurrent Tracking via Pyramid Recurrent Network and Optical Flow.
Ding MaWei BuXiangqian WuPublished in: BMVC (2018)
Keyphrases
- recurrent networks
- multiscale
- optical flow
- recurrent neural networks
- feature tracking
- dense optical flow
- optical flow estimation
- feed forward
- motion model
- lucas kanade
- coarse to fine
- motion segmentation
- scale space
- neural network
- biologically inspired
- point tracking
- image sequences
- image representation
- optical flow computation
- motion analysis
- optical flow algorithm
- moving camera
- motion field
- wavelet transform
- multiple scales
- particle filter
- image processing
- computer vision
- real time
- multiresolution
- d scene
- appearance model
- least squares
- object tracking
- smoothness constraint
- lucas kanade algorithm
- motion parameters
- flow field
- vector field
- edge detection
- artificial neural networks
- optic flow
- partial occlusion
- local binary pattern
- object detection
- decision making
- visual tracking
- challenging sequences
- qualitative and quantitative evaluation
- structure from motion
- back propagation