Recurrent Neural Network for (Un-)supervised Learning of Monocular VideoVisual Odometry and Depth.
Rui WangStephen M. PizerJan-Michael FrahmPublished in: CoRR (2019)
Keyphrases
- recurrent neural networks
- supervised learning
- visual odometry
- depth cues
- neural network
- depth images
- feed forward
- ego motion
- complex valued
- image sequences
- reservoir computing
- recurrent networks
- unsupervised learning
- depth map
- training data
- depth information
- learning tasks
- echo state networks
- relative depth
- neural model
- long short term memory
- depth perception
- feedforward neural networks
- single still image
- pose estimation
- semi supervised
- hidden layer
- active learning
- artificial neural networks
- training set
- reinforcement learning
- learning algorithm
- machine learning
- real time
- depth estimation
- autonomous navigation
- stereo vision
- stereo camera
- depth ordering
- nonlinear dynamic systems