---
title: Inverse Cooking
emoji: 👀
colorFrom: purple
colorTo: green
sdk: gradio
sdk_version: 5.32.0
app_file: app.py
pinned: false
license: mit
short_description: Generate recipes from food images using Facebook's model.
---


# Inverse Cooking - Recipe Generator from Images 🍽️

This Space deploys the [Inverse Cooking](https://github.com/facebookresearch/inversecooking) model developed by Facebook AI Research (FAIR).

Upload a food image and receive a predicted recipe with ingredients and cooking instructions.

> ⚠️ This model and code were originally developed by [facebookresearch/inversecooking](https://github.com/facebookresearch/inversecooking).  
> This deployment is for demonstration purposes only and does not claim ownership of the model or training data.

---

## 🔧 How it works

1. A Transformer-based encoder-decoder architecture takes an image of food.
2. It predicts a list of ingredients.
3. Based on the predicted ingredients, it generates cooking instructions.


![image/png](https://cdn-uploads.huggingface.co/production/uploads/65f8530280481173afd9eba2/2QrxZl98_ObPDfalE4oLm.png)

---

## 🧠 Model Info

- Encoder: ResNet-based visual encoder
- Decoder: Transformer decoder
- Trained on: Recipe1M+ dataset

---

## 📄 License

This deployment uses the original code and model weights licensed under **MIT License** from FacebookResearch.

---