MagicVO: End-to-End Monocular Visual Odometry through Deep Bi-directional Recurrent Convolutional Neural Network.
Jian JiaoJichao JiaoYaokai MoWeilun LiuZhongliang DengPublished in: CoRR (2018)
Keyphrases
- visual odometry
- end to end
- bi directional
- convolutional neural network
- autonomous navigation
- long range
- face detection
- ego motion
- kalman filtering
- depth images
- position information
- simultaneous localization and mapping
- multipath
- associative memory
- field of view
- real time
- dynamic environments
- optical flow
- neural network
- mobile robot
- robust estimation
- conditional random fields
- pairwise
- image sequences