https://github.com/arpanpramanik2003/fruit-veg-classification
This project is a deep learning-based image classification system that uses EfficientNet for accurate identification of fruits and vegetables from uploaded images. Built with TensorFlow and deployed using Streamlit, it provides a user-friendly interface, confidence scores, and visualization of predictions.
https://github.com/arpanpramanik2003/fruit-veg-classification
classification-model custom-cnn-model deep deployment efficientnet fruit-recognition image-classification keras pre-trained-model python3 sns streamlit-cloud streamlit-webapp ten transformers vegetable-cla
Last synced: 2 months ago
JSON representation
This project is a deep learning-based image classification system that uses EfficientNet for accurate identification of fruits and vegetables from uploaded images. Built with TensorFlow and deployed using Streamlit, it provides a user-friendly interface, confidence scores, and visualization of predictions.
- Host: GitHub
- URL: https://github.com/arpanpramanik2003/fruit-veg-classification
- Owner: arpanpramanik2003
- License: mit
- Created: 2025-02-11T18:32:14.000Z (over 1 year ago)
- Default Branch: master
- Last Pushed: 2025-02-16T15:29:42.000Z (over 1 year ago)
- Last Synced: 2025-05-13T22:28:36.253Z (about 1 year ago)
- Topics: classification-model, custom-cnn-model, deep, deployment, efficientnet, fruit-recognition, image-classification, keras, pre-trained-model, python3, sns, streamlit-cloud, streamlit-webapp, ten, transformers, vegetable-cla
- Language: Jupyter Notebook
- Homepage: https://fruit-veg-detection-custom-cnn-arpan.streamlit.app
- Size: 48.4 MB
- Stars: 1
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
[Dataset Link](https://www.kaggle.com/datasets/kritikseth/fruit-and-vegetable-image-recognition)
## 📌 Project Overview
This project is a **deep learning-based image classification system** that identifies different types of fruits and vegetables from an uploaded image. The model is built using **TensorFlow** and deployed using **Streamlit**. The current model uses **EfficientNet** for improved accuracy and performance.
## 🖥️ Tech Stack
- **Python** (Core programming language)
- **TensorFlow/Keras** (For model training and inference)
- **EfficientNet** (Pre-trained model for feature extraction and classification)
- **Streamlit** (For building the web UI)
- **NumPy & PIL** (For image preprocessing)
- **Matplotlib** (For visualization)
## 🎯 Features
- Upload an image of a fruit or vegetable
- Predict the class with confidence score
- Display confidence scores as a bar chart
- User-friendly and interactive interface
- Utilizes **EfficientNet** for high-accuracy predictions
## 🏗️ Project Structure
```
fruit_veg_classifier/
├── efficient_model.h5 # Trained EfficientNet model
│── app.py # Streamlit web app script
│── requirements.txt # Dependencies for the project
│── README.md # Project documentation
│── image.png # Project preview image
```
## 🚀 Installation & Usage
### 1️⃣ Clone the Repository
```sh
git clone https://github.com/arpanpramanik2003/fruit-veg-classification.git
cd fruit-veg-classification
```
### 2️⃣ Install Dependencies
```sh
pip install -r requirements.txt
```
### 3️⃣ Run the Streamlit App
```sh
streamlit run app.py
```
## 📷 Model & Image Preprocessing
- The model uses **EfficientNet** as the backbone for feature extraction.
- Input images are resized to **224x224 pixels** before inference.
- The model has achieved high accuracy during training and testing.
## 📊 Prediction Output
- **Class Label:** Name of the detected fruit/vegetable
- **Confidence Score:** Probability of prediction accuracy
- **Bar Chart:** Visualization of class probabilities
## 🛠️ Future Improvements
- Enhance accuracy further with data augmentation
- Deploy on reliable cloud platforms with minimal latency
- Add support for more categories and datasets
## 📜 License
This project is open-source and available under the **MIT License**.
---
📌 **Developed by Arpan Pramanik** | 💡 AI/ML Enthusiast 🚀