|
|
--- |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- en |
|
|
base_model: |
|
|
- google/siglip2-base-patch16-224 |
|
|
pipeline_tag: image-classification |
|
|
library_name: transformers |
|
|
tags: |
|
|
- brain |
|
|
- tumor |
|
|
- classification |
|
|
--- |
|
|
 |
|
|
|
|
|
# **BrainTumor-Classification-Mini** |
|
|
> **BrainTumor-Classification-Mini** is an image classification vision-language encoder model fine-tuned from **google/siglip2-base-patch16-224** for a single-label classification task. It is designed to classify brain tumor images using the **SiglipForImageClassification** architecture. |
|
|
|
|
|
|
|
|
```py |
|
|
Classification Report: |
|
|
precision recall f1-score support |
|
|
|
|
|
No Tumor 0.9975 0.9962 0.9969 1595 |
|
|
Glioma 0.9872 0.9947 0.9910 1321 |
|
|
Meningioma 0.9880 0.9821 0.9850 1339 |
|
|
Pituitary 0.9931 0.9931 0.9931 1457 |
|
|
|
|
|
accuracy 0.9918 5712 |
|
|
macro avg 0.9915 0.9915 0.9915 5712 |
|
|
weighted avg 0.9918 0.9918 0.9918 5712 |
|
|
``` |
|
|
|
|
|
 |
|
|
|
|
|
The model categorizes images into the following 4 classes: |
|
|
- **Class 0:** "No Tumor" |
|
|
- **Class 1:** "Glioma" |
|
|
- **Class 2:** "Meningioma" |
|
|
- **Class 3:** "Pituitary" |
|
|
|
|
|
# **Run with Transformers🤗** |
|
|
|
|
|
```python |
|
|
!pip install -q transformers torch pillow gradio |
|
|
``` |
|
|
|
|
|
```python |
|
|
import gradio as gr |
|
|
from transformers import AutoImageProcessor |
|
|
from transformers import SiglipForImageClassification |
|
|
from transformers.image_utils import load_image |
|
|
from PIL import Image |
|
|
import torch |
|
|
|
|
|
# Load model and processor |
|
|
model_name = "prithivMLmods/BrainTumor-Classification-Mini" |
|
|
model = SiglipForImageClassification.from_pretrained(model_name) |
|
|
processor = AutoImageProcessor.from_pretrained(model_name) |
|
|
|
|
|
def brain_tumor_classification(image): |
|
|
"""Predicts brain tumor category for an image.""" |
|
|
image = Image.fromarray(image).convert("RGB") |
|
|
inputs = processor(images=image, return_tensors="pt") |
|
|
|
|
|
with torch.no_grad(): |
|
|
outputs = model(**inputs) |
|
|
logits = outputs.logits |
|
|
probs = torch.nn.functional.softmax(logits, dim=1).squeeze().tolist() |
|
|
|
|
|
labels = { |
|
|
"0": "No Tumor", "1": "Glioma", "2": "Meningioma", "3": "Pituitary" |
|
|
} |
|
|
predictions = {labels[str(i)]: round(probs[i], 3) for i in range(len(probs))} |
|
|
|
|
|
return predictions |
|
|
|
|
|
# Create Gradio interface |
|
|
iface = gr.Interface( |
|
|
fn=brain_tumor_classification, |
|
|
inputs=gr.Image(type="numpy"), |
|
|
outputs=gr.Label(label="Prediction Scores"), |
|
|
title="Brain Tumor Classification", |
|
|
description="Upload an image to classify it into one of the 4 brain tumor categories." |
|
|
) |
|
|
|
|
|
# Launch the app |
|
|
if __name__ == "__main__": |
|
|
iface.launch() |
|
|
``` |
|
|
|
|
|
# **Intended Use:** |
|
|
|
|
|
The **BrainTumor-Classification-Mini** model is designed for brain tumor image classification. It helps categorize MRI images into predefined tumor types. Potential use cases include: |
|
|
|
|
|
- **Medical Diagnosis Assistance:** Supporting radiologists in preliminary tumor classification. |
|
|
- **AI-Assisted Healthcare:** Enhancing automated tumor detection in medical imaging. |
|
|
- **Research & Development:** Facilitating studies in AI-driven medical imaging solutions. |
|
|
- **Educational Purposes:** Helping students and professionals learn about tumor classification using AI. |
