Swin-MFA: A Multi-Modal Fusion Attention Network Based on Swin-Transformer for Low-Light Image Human Segmentation.
Xunpeng YiHaonan ZhangYibo WangShujiang GuoJingyi WuCien FanPublished in: Sensors (2022)
Keyphrases
- low light
- segmentation method
- image segmentation
- multiscale
- image analysis
- segmentation algorithm
- low light conditions
- image data
- edge detection
- pixel level
- test images
- visible spectrum
- input image
- motion blur
- pixel wise
- segmented images
- single image
- infrared images
- multi modal fusion
- noise level
- image structure
- feature points
- image noise
- similarity measure
- image matching
- object detection
- image edges
- high resolution
- object recognition
- image processing