Visual Domain Adaptation for Monocular Depth Estimation on Resource-Constrained Hardware.
Julia HornauerLazaros NalpantidisVasileios BelagiannisPublished in: CoRR (2021)
Keyphrases
- resource constrained
- domain adaptation
- depth estimation
- depth map
- stereo vision
- resource constraints
- embedded systems
- sensor networks
- wireless sensor networks
- stereo matching
- multiple sources
- cross domain
- semi supervised
- dynamic scenes
- low cost
- depth information
- scene understanding
- stereo pair
- labeled data
- semi supervised learning
- image sequences
- sentiment classification
- visual information
- super resolution
- disparity map
- stereo images
- test data
- multipath
- supervised learning
- target domain
- transfer learning
- pose estimation
- training data
- multi view
- input image
- active learning
- rfid tags
- high quality