Cook

Create your own dish by selecting its type and its ingedients!

What's on your plate?

Do you think this image is a real dish? (Leave us a feedback to skip the waiting time)

Kidding me? Not at all! Not so real Maybe, I can't tell It seems real It is so real!

Explore

Explore our existing gallery of dishes -- real versus synthetic!

Real Photo	Synthetic Image

Learn

Learn more about the AI behind CookGAN.

Motivation

Synthesizing photo-realistic food meal images from textual descriptions of its ingredients.

Methodology

The framework first builds an attention-based ingredients-image association model, which is then used to condition a generative neural network tasked with synthesizing meal images. Furthermore, a cycle-consistent constraint is added to further improve image quality and control appearance. Preliminary experiments show our model is able to generate meal image corresponding to the ingredients list.

Potential Applications

Beyond its ability to help understand the relationship between two domains (i.e. natural languange and vision), a well trained model also has other applications, to name a few, Text visualization, Data augmentation, Cooking skill learning.

SEQAM Lab Samsung AI Center
Department of Computer Science Cambridge, UK
Rutgers University
New Jersey, USA

Team
Fangda Han

Ricardo Guerrero

Vladimir Pavlovic

Guoyao Hao

Social Media

CookGAN

The AI for Cooking