|
|
--- |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- en |
|
|
base_model: |
|
|
- google/siglip2-base-patch16-224 |
|
|
pipeline_tag: image-classification |
|
|
library_name: transformers |
|
|
tags: |
|
|
- WBC |
|
|
- Type |
|
|
- Classifier |
|
|
--- |
|
|
 |
|
|
|
|
|
# **WBC-Type-Classifier** |
|
|
|
|
|
> **WBC-Type-Classifier** 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 different types of white blood cells (WBCs) using the **SiglipForImageClassification** architecture. |
|
|
|
|
|
```py |
|
|
Accuracy: 0.9891 |
|
|
F1 Score: 0.9893 |
|
|
|
|
|
Classification Report: |
|
|
precision recall f1-score support |
|
|
|
|
|
basophil 0.9822 0.9959 0.9890 1218 |
|
|
eosinophil 0.9994 0.9984 0.9989 3117 |
|
|
erythroblast 0.9835 0.9974 0.9904 1551 |
|
|
ig 0.9787 0.9693 0.9740 2895 |
|
|
lymphocyte 0.9893 0.9942 0.9918 1214 |
|
|
monocyte 0.9852 0.9852 0.9852 1420 |
|
|
neutrophil 0.9876 0.9838 0.9857 3329 |
|
|
platelet 1.0000 0.9996 0.9998 2348 |
|
|
|
|
|
accuracy 0.9891 17092 |
|
|
macro avg 0.9882 0.9905 0.9893 17092 |
|
|
weighted avg 0.9891 0.9891 0.9891 17092 |
|
|
``` |
|
|
|
|
|
 |
|
|
|
|
|
The model categorizes images into eight classes: |
|
|
- **Class 0:** "Basophil" |
|
|
- **Class 1:** "Eosinophil" |
|
|
- **Class 2:** "Erythroblast" |
|
|
- **Class 3:** "IG" |
|
|
- **Class 4:** "Lymphocyte" |
|
|
- **Class 5:** "Monocyte" |
|
|
- **Class 6:** "Neutrophil" |
|
|
- **Class 7:** "Platelet" |
|
|
|
|
|
# **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/WBC-Type-Classifier" |
|
|
model = SiglipForImageClassification.from_pretrained(model_name) |
|
|
processor = AutoImageProcessor.from_pretrained(model_name) |
|
|
|
|
|
def wbc_classification(image): |
|
|
"""Predicts WBC type for a given blood cell 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": "Basophil", "1": "Eosinophil", "2": "Erythroblast", "3": "IG", |
|
|
"4": "Lymphocyte", "5": "Monocyte", "6": "Neutrophil", "7": "Platelet" |
|
|
} |
|
|
predictions = {labels[str(i)]: round(probs[i], 3) for i in range(len(probs))} |
|
|
|
|
|
return predictions |
|
|
|
|
|
# Create Gradio interface |
|
|
iface = gr.Interface( |
|
|
fn=wbc_classification, |
|
|
inputs=gr.Image(type="numpy"), |
|
|
outputs=gr.Label(label="Prediction Scores"), |
|
|
title="WBC Type Classification", |
|
|
description="Upload a blood cell image to classify its WBC type." |
|
|
) |
|
|
|
|
|
# Launch the app |
|
|
if __name__ == "__main__": |
|
|
iface.launch() |
|
|
``` |
|
|
|
|
|
# **Intended Use:** |
|
|
|
|
|
The **WBC-Type-Classifier** model is designed to classify different types of white blood cells from blood smear images. Potential use cases include: |
|
|
|
|
|
- **Medical Diagnostics:** Assisting pathologists in identifying different WBC types for diagnosis. |
|
|
- **Hematology Research:** Supporting studies related to blood cell morphology and disease detection. |
|
|
- **Automated Blood Analysis:** Enhancing automated diagnostic tools for rapid blood cell classification. |
|
|
- **Educational Purposes:** Providing insights and training data for medical students and researchers. |
