You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

CerS-Path

Weight: weights/CerS-Path.pth
Input: RGB pathology patches
Output: 1024-dimensional features

Install

python -m pip install -r requirements.txt

Preprocessing

transforms.Compose([
    transforms.Resize(
        224,
        interpolation=InterpolationMode.BILINEAR,
        antialias=True,
    ),
    transforms.ToTensor(),
    transforms.Normalize(
        mean=(0.485, 0.456, 0.406),
        std=(0.229, 0.224, 0.225),
    ),
])

No random augmentation or stain normalization is used during feature extraction. The reported evaluation used 20X, 256 × 256 JPG patches.

Usage

python extract_features.py \
  --input /path/to/patches \
  --output features.pt

Patch filenames should follow x_y.jpg. The output tensor has shape [number_of_patches, 1024].

Downloads last month: -