| --- |
| language: en |
| license: mit |
| tags: |
| - deep-learning |
| - cancer-detection |
| - histopathology |
| - tensorflow |
| - efficientnet |
| - vision-transformer |
| - ViT |
| - medical-imaging |
| model_name: EfficientNetV2S & ViT-Hybrid for Histopathologic Cancer Detection |
| library_name: tensorflow |
| datasets: |
| - histopathologic-cancer-detection |
| - PatchCamelyon |
| pipeline_tag: image-classification |
| --- |
| |
|
|
| # Histopathologic Cancer Detection - EfficientNetV2S & ViT-Hybrid |
| This repository contains models for detecting metastatic cancer in histopathologic images. |
| - **EfficientNetV2S**: A Baseline CNN-based model for local feature extraction. |
| - **ViT-Hybrid**: A Transformer-based model that learns global dependencies. |
| Both models were trained on the [Histopathologic Cancer Detection Kaggle dataset](https://www.kaggle.com/competitions/histopathologic-cancer-detection/data) |
|
|
| ## Model Performance |
| - **EfficientNetV2S** |
| - Accuracy: 93.59% (Private), 93.74% (Public) |
| - AUC: 0.9774 |
| - **ViT-Hybrid** |
| - Accuracy: 95.07% (Private), 94.87% (Public) |
| - AUC: 0.9791 |
| - **ViT-Hybrid + TTA (Test-Time Augmentation)** |
| - Accuracy: 96.50% (Private), 96.75% (Public) |
|
|
| ## Model Use |
|
|
| ```sh |
| from huggingface_hub import hf_hub_download |
| from tensorflow.keras.models import load_model |
| ``` |
|
|
| # Download EfficientNetV2S model |
| ```sh |
| model_path = hf_hub_download(repo_id="MooseML/EfficientNet-Cancer-Detection", filename="efficientnet_cancer_model.h5") |
| model = load_model(model_path) |
| ``` |
|
|
| # Download ViT-Hybrid model |
| ```sh |
| model_path_vit = hf_hub_download(repo_id="MooseML/EfficientNet-Cancer-Detection", filename="ViT_hybrid_cancer_model.h5") |
| model_vit = load_model(model_path_vit) |
| ``` |
|
|
| ## Github and Kaggle Links for Full Training Pipeline |
| - Full Training Code: [GitHub Repository](https://github.com/MooseML/Histopathologic-Cancer-Detection) |
| - Kaggle Competition: [Histopathologic Cancer Detection](https://www.kaggle.com/competitions/histopathologic-cancer-detection) |
|
|
|
|
| --- |
| license: mit |
| --- |
