| --- |
| |
| |
| license: cc-by-nc-4.0 |
| datasets: |
| - TLAIM/TAIX-Ray |
| language: |
| - en |
| tags: |
| - medical |
| - x-ray |
| - radiograph |
| - thorax |
| --- |
| |
| # TAIX-Ray Models |
|
|
| This repository provides two trained deep learning models for classifying X-ray images from the TAIX-Ray dataset: |
|
|
| 1. Binary Classification Model - Classifies X-ray images into two categories (normal vs. abnormal). |
|
|
| 2. Ordinal Classification Model - Predicts severity levels based on ordinal categories. |
|
|
|
|
| <br> |
|
|
| ## Code & Details: |
| Model definitions, training, and evaluation code are available at https://github.com/mueller-franzes/TAIX-Ray |
|
|
| ## How to Use |
|
|
| ### Prerequisites |
| Ensure you have the following dependencies installed: |
|
|
| ```bash |
| pip install huggingface_hub |
| ``` |
|
|
|
|
| ### Download |
|
|
| ```python |
| from huggingface_hub import hf_hub_download |
| |
| |
| # Download the checkpoint file from Hugging Face Hub |
| file_path = hf_hub_download( |
| repo_id="TLAIM/TAIX-Ray", |
| filename="binary.ckpt", # binary.ckpt or ordinal.ckpt |
| ) |
| |
| # Check if the file has been correctly downloaded |
| print(f"Checkpoint downloaded to: {file_path}") |
| ``` |
|
|
|
|
|
|
