DeepVO: Towards end-to-end visual odometry with deep Recurrent Convolutional Neural Networks.
Sen WangRonald ClarkHongkai WenNiki TrigoniPublished in: ICRA (2017)
Keyphrases
- end to end
- visual odometry
- convolutional neural networks
- autonomous navigation
- long range
- ego motion
- depth images
- position information
- ad hoc networks
- kalman filtering
- multipath
- simultaneous localization and mapping
- range data
- congestion control
- camera pose
- field of view
- camera calibration
- dynamic environments
- computer vision
- wifi
- d objects
- mobile robot
- optical flow