Charge Own Job: Saliency Map and Visual Word Encoder for Image-Level Semantic Segmentation.
Yuhui GuoXun LiangHui TangXiangping ZhengBo WuXuan ZhangPublished in: ECML/PKDD (3) (2022)
Keyphrases
- image classification
- visual words
- input image
- scene classification
- saliency map
- image features
- image representation
- image retrieval
- keypoints
- pixel level
- bag of words
- natural scenes
- bag of features
- multiscale
- image data
- image content
- spatial information
- segmentation method
- image segmentation
- object categories
- superpixels
- high resolution
- visual attention
- object recognition
- low level
- image collections
- similarity measure
- test images
- video frames
- natural images
- image database
- co occurrence
- object detection
- feature vectors
- information extraction
- probabilistic model