FoodVision empowers you to see your food, and know your food with lightning-fast image classification. Leveraging the powerful EfficientNetB3 model, it identifies 101 diverse food categories instantly.
See FoodVision in real-time on Hugging Face Spaces:
https://huggingface.co/spaces/Anithprakash/FOOD_EUROPLATE
An EfficientNetB3 model was employed for classifying food images. This model excels in balancing accuracy and computational efficiency. A pre-existing dataset named "Food101" was utilized for training. While the model achieved an accuracy score exceeding 65%, it's crucial to acknowledge limitations. Food101 primarily focuses on food items commonly found in North America, potentially leading to a bias towards this region's cuisine. This bias might affect the model's ability to accurately classify food images from other parts of the world, particularly Europe.
Contributions are welcome! If you find issues or want to add new features, please submit a pull request.
Symptoms: Users are unable to access the web app, encountering errors or connection issues.
Possible Solutions:
- Check Internet Connection.
- Ensure browser compatibility.
- Verify firewall or network settings.
For inquiries, collaboration opportunities, or feedback, please reach out to me via email: anithprakashan321@gmail.com