Recurrent convolutional neural networks for object-class segmentation of RGB-D video.
Mircea Serban PavelHannes SchulzSven BehnkePublished in: IJCNN (2015)
Keyphrases
- object class
- convolutional neural networks
- object detection
- bounding box
- object recognition
- object segmentation
- object classes
- viewpoint
- object categories
- superpixels
- segmentation method
- shape prior
- segmentation algorithm
- image segmentation
- partial occlusion
- appearance model
- level set
- multiscale
- image patches
- computer vision
- color images
- deformable models
- object tracking
- target object
- visual tracking
- test images
- energy function
- face detection
- natural images
- three dimensional
- image processing