TreeFormer: a Semi-Supervised Transformer-based Framework for Tree Counting from a Single High Resolution Image.
Hamed Amini AmirkolaeeMiaojing ShiMark MulliganPublished in: CoRR (2023)
Keyphrases
- high resolution
- semi supervised
- input image
- image analysis
- test images
- image segmentation
- high resolution images
- image data
- satellite images
- image features
- low resolution
- image processing
- multiscale
- image classification
- bayesian framework
- single image
- segmentation method
- image content
- low resolution images
- image generation
- lower resolution
- feature points
- high frequency
- super resolution
- high quality
- remote sensing
- field of view
- pixel values
- dimensionality reduction
- active learning
- decision trees