Learning Dense Visual Descriptors using Image Augmentations for Robot Manipulation Tasks.
Christian GrafDavid B. AdrianJoshua WeilMiroslav GabrielPhilipp SchillingerMarkus SpiesHeiko NeumannAndras G. KupcsikPublished in: CoRR (2022)
Keyphrases
- manipulation tasks
- visual descriptors
- image data
- autonomous robots
- image retrieval
- robot navigation
- image content
- learning algorithm
- image representation
- image features
- image classification
- input image
- robotic systems
- similarity measure
- feature points
- visual features
- transfer learning
- spatial relationships
- motion planning
- position and orientation
- low level
- reinforcement learning