Multimodal Few-Shot Object Detection with Meta-Learning Based Cross-Modal Prompting.
Guangxing HanJiawei MaShiyuan HuangLong ChenRama ChellappaShih-Fu ChangPublished in: CoRR (2022)
Keyphrases
- cross modal
- meta learning
- object detection
- multi modal
- learning tasks
- inductive learning
- model selection
- multimedia retrieval
- computer vision
- machine learning algorithms
- object categories
- object recognition
- multi class
- video sequences
- visual recognition
- feature selection
- decision trees
- data mining
- machine learning
- visual similarity
- base classifiers
- multimedia databases
- database
- visual data
- visual features
- knowledge base
- video data
- image retrieval
- video content
- data sets
- key frames
- training samples
- knowledge representation
- low level
- active learning
- pairwise
- feature space
- high level
- learning algorithm