Multimodal Hateful Memes Detection via Image Caption Supervision.
Huaicheng FangFuqing ZhuJizhong HanSonglin HuPublished in: SmartWorld/UIC/ScalCom/DigitalTwin/PriComp/Meta (2022)
Keyphrases
- image content
- image features
- bounding box
- image retrieval
- multiscale
- image data
- input image
- image classification
- image analysis
- image representation
- region of interest
- feature points
- image pixels
- image collections
- template matching
- image segmentation
- hough transform
- image regions
- detection algorithm
- multi modal
- low level
- segmentation method
- test images
- edge detection
- caption text
- single image
- visual features
- high resolution
- image matching
- lighting conditions
- active learning
- computer vision