A team of researchers from Tel Aviv University has developed a neural network able to read a recipe and generate a picture what the finished product would look like and cooked. As if DeepFakes It was not serious enough, we can not be sure that the delicious dishes we see online are real.
The Tel-Aviv team, composed of researchers Ori Bar El, Ori Licht and Netanel Yosephian created their artificial intelligence using a modified version of a generative contradictory network (GAN) called StackGAN V2 and the gigantic recipe1M 52K image / recipe combinations dataset.
Basically, the team has developed an artificial intelligence that can take almost any list of ingredients and instructions and determine the appearance of the finished food product.
Researcher Ori Bar El told TNW:
[It] it all started when I asked my grandmother for a recipe for her legendary fish chops with tomato sauce. Due to her advanced age, she did not remember the exact recipe. So, I was wondering if I could build a system that would produce the recipe from an image of the food. After thinking about this task for a moment, I concluded that it was too difficult for a system to get an exact recipe with real amounts and with "hidden" ingredients such as salt. , pepper, butter, flour, etc.
Then I wondered if I could do the opposite instead. Namely, generate food images from the recipes. We think this task is very difficult for humans to do, let alone for computers. As most current AI systems are trying to replace human experts in human-friendly tasks, we thought it would be interesting to solve a type of task that goes beyond the capabilities of humans. As you can see, this can be done to a certain extent.
Researchers also recognize, in their white paper, that the system is not quite perfect yet:
Note that the image quality of the recipe1M dataset is poor compared to the images in the CUB and Oxford102 datasets. This results in many blurry images with poor lighting conditions, "porridge-like images" and the fact that the images are not square-shaped (which makes it difficult to model). This may explain the fact that both models have managed to generate images of "porridge-like" foods (eg, pasta, rice, soups, salads), but struggle to generate food-shaped images. distinctive (eg hamburger, chicken, beverages). ).
To our knowledge, this is the only artificial intelligence of this type. So do not expect it to be an application on your phone anytime soon. But the writing is on the wall. And, if it's a recipe, the artificial intelligence of the Tel Aviv team can turn it into an image that looks good enough that, according to the research paper, humans sometimes prefer it. a picture of reality.
What do you think?
The team intends to continue developing the system, hopefully in areas other than food. Ori Bar El told us:
We plan to extend the work by training our system on the rest of the recipes (we have about 350,000 additional images), but the problem is that the current dataset is of poor quality. We did not find any other datasets available for our needs, but we could create a dataset containing the text for children's books and the corresponding images.
These talented researchers may have damned gourmets Instagram in a world where we can not really know for sure if what's drooling is real, or the vision of a robot's breath. "
This is probably the right time for all of us to go to the real world and stick our face to real food. You know, the kind created by scientists and prepared by robots.
Real World Photoshop: Procter & Gamble debuted with a handheld device that could replace makeup