Weakly Supervised Deep Convolutional Networks for Fine-Grained Object Recognition in Multispectral Images.
Bulut AygünesSelim AksoyRamazan Gökberk CinbisPublished in: IGARSS (2019)
Keyphrases
- fine grained
- weakly supervised
- deep learning
- multispectral images
- object class
- object recognition
- multispectral
- coarse grained
- spatial resolution
- topic models
- object detection
- superpixels
- natural images
- object detectors
- object categories
- access control
- unsupervised learning
- relation extraction
- d objects
- remote sensing images
- computer vision
- viewpoint
- semi supervised
- image analysis
- object classes
- remote sensing
- fusion method
- edge information
- color images
- information retrieval
- appearance model
- named entities
- image processing
- image representation
- image features
- multiscale
- image sequences
- high level
- image segmentation