FA-Net: A Fused Feature for Multi-Head Attention Recoding Network for Pear Leaf Nutritional Deficiency Diagnosis with Visual RGB-Image Depth and Shallow Features.
Yi SongLi LiuYuan RaoXiaodan ZhangXiu JinPublished in: Sensors (2023)
Keyphrases
- image features
- low level
- feature values
- global features
- invariant features
- feature detectors
- individual features
- multiple features
- image classification
- feature representation
- geometrical features
- feature vectors
- visual appearance
- image similarity
- image representation
- image retrieval
- color bands
- extracted features
- feature descriptors
- input image
- feature set
- visual features
- object features
- visual patterns
- textural features
- low level visual features
- image content
- visual attributes
- visual information
- image matching
- spatial information
- image regions
- visual cues
- spatial layout
- multiscale
- feature detection
- feature extractor
- hsv color
- test images
- single image
- visual concepts
- depth data
- keypoints
- visual field
- depth map
- single feature
- object recognition
- high resolution
- image segmentation
- feature extraction
- feature space
- feature fusion
- color images
- segmented regions
- gray scale
- color space