Sample Select Policy

Which images to label for few-shot medical landmark detection?

Environment

python == 3.5/3.6, 
pytorch >= 1.1.0, 
torchvison >= 0.6

Data preparation

We train/test our model on Cephalometric Dataset

We expect the directory structure to be the following:

path/to/cephalometric
	400_junior
		001.txt
		...
	400_senior
		001.txt
		...
	RawImage
		TrainingData
			001.bmp
			...
		Test1Data
			151.bmp
			...
		Test2Data
			301.bmp
			...

Steps

1. Train the feature extractor

python -m sc.ssl.ssl --tag run

2. extract SIFT key points

python -m sc.select.sift_select --tag sift

3. calculate similarities (Respective score)

python -m sc.select.maxsim_sift --tag sift

4. select templates

python -m sc.select.selct_ids --tag sift 

Evaluation

1. Estimate all MRE

python -m sc.select.test_by_multi --tag xx

2. Test templates

python -m sc.select.test_specific_ids --indices xxx

License

This code is released under the Apache 2.0 license. Please see the LICENSE file for more information.

Citation

@InProceedings{Quan_2022_CVPR,
    author    = {Quan, Quan and Yao, Qingsong and Li, Jun and Zhou, S. Kevin},
    title     = {Which Images To Label for Few-Shot Medical Landmark Detection?},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2022},
    pages     = {20606-20616}
}