Depth Not Needed - An Evaluation of RGB-D Feature Encodings for Off-Road Scene Understanding by Convolutional Neural Network.
Christopher J. HolderToby P. BreckonXiong WeiPublished in: CoRR (2018)
Keyphrases
- scene understanding
- convolutional neural network
- object detection
- object recognition
- d scene
- object hypotheses
- indoor scenes
- face detection
- robot navigation
- vision system
- scene recognition
- video surveillance
- scene categorization
- depth images
- scene labeling
- depth information
- depth map
- support vector machine
- scene interpretation
- feature selection
- geometric reasoning
- appearance model
- single image
- image features
- depth data
- feature vectors
- depth cameras
- high quality
- microsoft kinect
- neural network
- real time