S-MAT: Semantic-Driven Masked Attention Transformer for Multi-Label Aerial Image Classification.
Hongjun WuCheng XuHongzhe LiuPublished in: Sensors (2022)
Keyphrases
- multi label
- image classification
- multi label classification
- image features
- image annotation
- multi instance
- image representation
- visual features
- bag of words
- binary classification
- feature extraction
- graph cuts
- multi label learning
- natural language
- text categorization
- hierarchical text categorization
- semantic information
- protein function prediction
- multiple labels
- low level features
- svm classifier
- cost sensitive
- visual words
- neural network
- knowledge discovery
- image retrieval
- similarity measure
- high level