InterpoNet, A brain inspired neural network for optical flow dense interpolation.
Shay ZweigLior WolfPublished in: CoRR (2016)
Keyphrases
- optical flow
- neural network
- dense optical flow
- optical flow fields
- scene flow
- optical flow computation
- motion field estimation
- optical flow field
- image sequences
- optical flow estimation
- dense correspondence
- flow field
- least squares
- point tracking
- d scene
- artificial neural networks
- back propagation
- motion field
- flow estimation
- human brain
- dense motion estimation
- neural network is trained
- qualitative and quantitative evaluation
- image interpolation
- motion estimation
- motion model
- brain images
- recurrent neural networks
- interpolation method
- linear interpolation
- fuzzy logic
- computer vision
- brain activity
- feature tracking
- vector field
- dense optical flow estimation
- displacement field
- stereo correspondence
- velocity field
- motion analysis
- neural network model
- fault diagnosis
- moving objects
- three dimensional
- genetic algorithm