Recurrent Neural Network for (Un-)Supervised Learning of Monocular Video Visual Odometry and Depth.
Rui WangStephen M. PizerJan-Michael FrahmPublished in: CVPR (2019)
Keyphrases
- visual odometry
- recurrent neural networks
- depth images
- ego motion
- autonomous navigation
- long range
- dynamic scenes
- feed forward
- real time
- video sequences
- video data
- neural network
- kalman filtering
- moving camera
- field of view
- pose estimation
- optical flow
- reinforcement learning
- hidden layer
- depth map
- position information
- d scene
- image sequences
- artificial neural networks
- depth information
- simultaneous localization and mapping
- camera motion
- training data
- stereo camera
- robust estimation
- low resolution
- multi camera
- semi supervised
- kalman filter
- camera pose
- flow field
- mobile robot
- space time
- computer vision
- multi view
- dynamic environments
- learning algorithm