Pre-computed Liquid Spaces with Generative Neural Networks and Optical Flow.
Boris BonevLukas PrantlNils ThuereyPublished in: CoRR (2017)
Keyphrases
- pre computed
- optical flow
- neural network
- generative model
- pattern recognition
- optical flow estimation
- motion field
- image sequences
- least squares
- d scene
- genetic algorithm
- flow field
- motion model
- data cube
- feed forward
- unsupervised learning
- motion estimation
- optical flow computation
- velocity field
- feature tracking
- neural network model
- index tree
- dense optical flow
- disk space
- camera motion
- moving objects
- artificial neural networks
- fuzzy logic
- mathematical model
- database
- qualitative and quantitative evaluation
- probabilistic model
- multilayer perceptron
- boundary conditions
- data warehouse
- nearest neighbor
- multi dimensional
- management system
- motion segmentation
- category specific
- similarity search
- back propagation
- fault diagnosis