Welcome to Gourmet Vision, where artificial intelligence meets culinary expertise! This project harnesses the power of YOLOv8 to create a sophisticated food detection model capable of identifying thirty six food items
- 🧠 Utilizes state-of-the-art YOLOv8 architecture
- 🍽️ Specialized in detecting Italian cuisine favorites
- 🚀 Easy-to-use training script for customization
- 📊 Comprehensive dataset with train, validation, and test splits
Our kitchen of code is organized as follows:
data.yaml: The recipe book - configuration file for dataset paths and class namestrain.py: The master chef - Python script that trains our AI food critic
Our AI is trained on a smorgasbord of images, carefully curated and divided:
- 🍳 Train set:
.../train/images - 🥘 Validation set:
.../valid/images - 🍲 Test set:
.../test/images
On today's menu, we're serving thirty six delicious treats:
- 🥤 AW cola
- 🥩🇨🇳 Beijing Beef
- 🍜 Chow Mein
- 🍚 Fried Rice
- 🥔 Hashbrown
- 🍤🥜 Honey Walnut Shrimp
- 🐔🌶️ Kung Pao Chicken
- 🐔🥬 String Bean Chicken Breast
- 🥬 Super Greens
- 🍗🍊 The Original Orange Chicken
- 🍚 White Steamed Rice
- 🍚🌶️ Black pepper rice bowl
- 🍔 Burger
- 🥕🥚 Carrot eggs
- 🍔🧀 Cheese burger
- 🍗🧇 Chicken waffle
- 🍗 Chicken nuggets
- 🥬 Chinese cabbage
- 🌭🇨🇳 Chinese sausage
- 🌽 Crispy corn
- 🍛 Curry
- 🍟 French fries
- 🍗 Fried chicken
- 🍗 Fried chicken
- 🥟 Fried dumplings
- 🍳 Fried eggs
- 🥭🐔 Mango chicken pocket
- 🍔🧀 Mozza burger
- 🌱 Mung bean sprouts
- 🍗 Nugget
- 🥔 Perkedel
- 🍚 Rice
- 🥤 Sprite
- 🧀 Tostitos cheese dip sauce
- 🥔 Triangle hash brown
- 🥬 Water spinach
Our model undergoes rigorous training to become the Gordon Ramsay of food detection:
- 🏋️♂️ Training regimen: 250 epochs
- 🍱 Batch size: 16 images at a time
- 👁️ Image size: A crisp 640x640 resolution
- 💪 Hardware: GPU-powered (RTX 3060 Laptop) for lightning-fast learning
- 🛑 Early stopping: We know when the dish is perfectly cooked (patience of 20)
To train your very own AI food critic:
- Ensure your kitchen (development environment) is stocked with the necessary ingredients (dependencies):
pip install ultralytics- Fire up the stove (run the training script):
python train.py- Sit back and watch as your AI chef learns the art of food detection. The fruits of its labor (trained model and results) will be plated up in
.../food-model.
Our gourmet dataset is sourced from the prestigious Roboflow kitchen:
- 👨🍳 Head Chef (Workspace): kaavin-study-drive
- 🍽️ Signature Dish (Project): food-k8grc
- 🥇 Michelin Stars (Version): 1
- 📜 Recipe Sharing Policy (License): MIT
- 🌐 Reservations (URL): Book a table at our Roboflow restaurant
This project is served under the MIT License. Feel free to savor, modify, and share the recipe, but don't forget to credit the original chefs!
We hope you enjoy using Gourmet Vision as much as we enjoyed creating it. May your code be bug-free and your detections accurate!
Remember, in the world of AI and food, pixels are the new flavors, and neural networks are the new taste buds. Happy coding and happy eating! 🎉👨🍳🤖