Unsupervised Vision-and-Language Pre-training Without Parallel Images and Captions.
Liunian Harold LiHaoxuan YouZhecan WangAlireza ZareianShih-Fu ChangKai-Wei ChangPublished in: NAACL-HLT (2021)
Keyphrases
- image database
- image data
- image features
- image retrieval
- image classification
- ground truth
- image analysis
- three dimensional
- input image
- image collections
- image understanding
- edge detection
- test images
- object recognition
- similarity measure
- computer vision
- neural network
- segmentation algorithm
- segmentation method
- image annotation
- image search
- unsupervised learning
- programming language
- semi supervised
- training set
- visual features
- feature points
- supervised learning
- region of interest
- lighting conditions
- natural language
- real time
- supervised training