Found a Reason for me? Weakly-supervised Grounded Visual Question Answering using Capsules.
Aisha Urooj KhanHilde KuehneKevin DuarteChuang GanNiels da Vitoria LoboMubarak ShahPublished in: CoRR (2021)
Keyphrases
- question answering
- weakly supervised
- relation extraction
- named entities
- information extraction
- information retrieval
- topic models
- natural language processing
- object class
- natural language
- visual information
- semi supervised
- automatic extraction
- visual features
- named entity recognition
- superpixels
- domain specific
- machine learning
- low level
- image segmentation
- object recognition
- knowledge discovery
- graphical models
- object detection
- ground truth