Self-Supervised Image-to-Point Distillation via Semantically Tolerant Contrastive Loss.
Anas MahmoudJordan S. K. HuTianshu KuaiAli HarakehLiam PaullSteven L. WaslanderPublished in: CoRR (2023)
Keyphrases
- single image
- input image
- image content
- multiscale
- image classification
- pose determination
- image data
- low level
- image features
- image segmentation
- high resolution
- edge detection
- image retrieval
- image matching
- template matching
- test images
- image analysis
- grey level
- image regions
- vector field
- lighting conditions
- hough transform
- pixel values
- similarity measure
- segmentation method
- color images