Fine-tuning Monocular Depth-Estimator Artificial Neural Networks Trained on Synthetic RGB-D Data Sets for Real Scenes.
Seyed Erfan Seyed RoghaniEmre KoyuncuPublished in: CoDIT (2020)
Keyphrases
- real scenes
- fine tuning
- artificial neural networks
- data sets
- depth map
- depth information
- depth cues
- depth data
- depth estimation
- image sequences
- depth images
- training set
- augmented reality
- pose estimation
- stereo images
- multi view
- neural network
- fine tuned
- d scene
- scene geometry
- stereo pair
- ego motion
- stereo matching
- input image
- training data
- low resolution
- input data
- motion estimation
- depth from defocus