Unsupervised Learning of Depth and Deep Representation for Visual Odometry from Monocular Videos in a Metric Space.
Xiaochuan YinChengju LiuPublished in: CoRR (2019)
Keyphrases
- metric space
- visual odometry
- ego motion
- depth images
- similarity search
- autonomous navigation
- kalman filtering
- long range
- distance function
- range queries
- position information
- image sequences
- optical flow
- depth information
- mobile robot
- depth map
- high dimensional
- moving camera
- dynamic scenes
- computer vision
- pose estimation
- field of view
- data sets
- kalman filter
- vector space
- multi dimensional
- image data
- range data
- stereo camera
- image processing
- simultaneous localization and mapping