Modelpublic

mziarehman4353/cell-segmentation-bbbc038

sourceHugging Facemitupdated 5mo agoView on Hugging Face

0likes

Model Card

🧬 Cell Segmentation & Classification (BBBC038)

This model performs:

—🔹 Cell segmentation using DeepLabV3+
—🔹 Cell classification (small vs large cells)

📊 Results

—Dice Score: 0.85
—IoU Score: 0.74
—Classification Accuracy: 91%

🧠 Model

—Backbone: ResNet34
—Architecture: DeepLabV3+

🚀 Usage

python

import torch
import segmentation_models_pytorch as smp

model = smp.DeepLabV3Plus(
    encoder_name="resnet34",
    encoder_weights=None,
    classes=1
)

model.load_state_dict(torch.load("deeplabv3plus.pth"))
model.eval()
📁 Dataset
BBBC038 (Broad Bioimage Benchmark Collection)
👨‍💻 Author

Zia Ul Rehman Zafar