DeepVIO is a self-supervised deep network for monocular VIO that projects stereo-derived 3D geometric constraints to train 2D optical flow and IMU fusion networks, outperforming prior learned methods on KITTI and EuRoC.
Unsupervised learning of depth and ego-motion from video
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2019 2verdicts
UNVERDICTED 2representative citing papers
Multi-step predictive generative models form stable belief states capturing environment layout and agent pose, yielding higher data efficiency on RL tasks than model-free agents.
citing papers explorer
-
DeepVIO: Self-supervised Deep Learning of Monocular Visual Inertial Odometry using 3D Geometric Constraints
DeepVIO is a self-supervised deep network for monocular VIO that projects stereo-derived 3D geometric constraints to train 2D optical flow and IMU fusion networks, outperforming prior learned methods on KITTI and EuRoC.
-
Shaping Belief States with Generative Environment Models for RL
Multi-step predictive generative models form stable belief states capturing environment layout and agent pose, yielding higher data efficiency on RL tasks than model-free agents.