A self-supervised scheme uses a vision model as a teacher to train a neural drone model from onboard data, improving velocity estimates and VIO accuracy at high speeds, with a proposed occlusion-handling loss that cuts pose RMSE by 15%.
Visual-inertial navigation: A concise review,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Self-Supervised Monocular Visual Drone Model Identification through Improved Occlusion Handling
A self-supervised scheme uses a vision model as a teacher to train a neural drone model from onboard data, improving velocity estimates and VIO accuracy at high speeds, with a proposed occlusion-handling loss that cuts pose RMSE by 15%.