link of b end and streemleet f end :https://huggingface.co/spaces/jarvisemitra/braintumer

## 🔁 Model Pipeline MRI Image ↓ Preprocessing (Resize + Normalize) ↓ Data Augmentation ↓ CNN Model (ResNet / EfficientNet) ↓ Feature Extraction ↓ Fully Connected Layer ↓ Softmax Output ↓ Prediction + Visualization ## 📊 Sample Prediction

## 🔥 Abnormality Detection (Heatmap)

## 🚀 Features

✔ Brain tumor classification (multi-class)
✔ Deep learning with transfer learning
✔ Image preprocessing + augmentation
✔ Heatmap-based abnormal region detection
✔ Streamlit UI for real-time prediction

🧠 How It Works

The system takes MRI images as input, preprocesses them, and passes them through a pretrained CNN model. The model extracts features and predicts the tumor type using a softmax classifier. Additionally, abnormal regions are highlighted using heatmap-based visualization.