Incorporating Category Taxonomy in Deep Reinforcement Learning Based Image Hashing.
Qiang FuLinsen DongZiyuan LiuYong LuoYonggang WenYing LiLing-Yu DuanPublished in: ICME (2019)
Keyphrases
- reinforcement learning
- single image
- image data
- input image
- multiscale
- image content
- low level
- edge detection
- image collections
- segmentation method
- image retrieval
- image analysis
- vector field
- image features
- image segmentation
- test images
- region of interest
- feature points
- template matching
- image classification
- keypoints
- high resolution
- image set
- improve recognition accuracy
- image quality
- hierarchical structure
- image representation
- medical images
- d objects
- similarity measure