AI to Create Pizza Receipts

MIT researchers used AI to create new pizza receipts; some of them are really interesting
12 September 2018   841

Researchers at MIT conducted an experiment to create recipes for pizza by artificial intelligence. Working within the framework of the project How To Generate (Almost) Anything, scientists used a recurrent neural network with open source called textgenrnn. The training was conducted on hundreds of author's recipes from culinary blogs.

To assess the taste qualities of the generated recipes, the researchers turned to the culinary specialists for help. Pizzeria Crush Pizza in Boston, Massachusetts, agreed to help implement artificial intelligence.

The local chef noted that in some recipes there are no key ingredients of the dish - meat topping, sauce or cheese. In addition, some components are quite difficult to find. In the pizzerias there was no "crushed caramel cheese" or "farmer's filling from a walnut".

As a result of oral testing of prototypes (in other words, having tried a piece), scientists came to the conclusion that some recipes turned out to be rather not bad. The top includes the following pizza recipes:

  • blueberries, spinach and feta cheese;
  • bacon, avocado and peach;
  • apricot, pear, cranberry and ricotta;
  • sweet potatoes, beans and brie;
  • shrimp, jam and assorted Italian sausages.

Pizza Receipts by AI
Pizza Receipts by AI

The last recipe for pizza was so successful that the chef of the pizzeria promised to think about including it in the main menu.

Nvidia to Open SPADE Source Code

SPADE machine learning system creates realistic landscapes based on rough human sketches
15 April 2019   671

NVIDIA has released the source code for the SPADE machine learning system (GauGAN), which allows for the synthesis of realistic landscapes based on rough sketches, as well as training models associated with the project. The system was demonstrated in March at the GTC 2019 conference, but the code was published only yesterday. The developments are open under the non-free license CC BY-NC-SA 4.0 (Creative Commons Attribution-NonCommercial-ShareAlike 4.0), allowing use only for non-commercial purposes. The code is written in Python using the PyTorch framework.

Sketches are drawn up in the form of a segmented map that determines the placement of exemplary objects on the scene. The nature of the generated objects is set using color labels. For example, a blue fill turns into sky, blue into water, dark green into trees, light green into grass, light brown into stones, dark brown into mountains, gray into snow, a brown line into a road, and a blue line into the river. Additionally, based on the choice of reference images, the overall style of the composition and the time of day are determined. The proposed tool for creating virtual worlds can be useful to a wide range of specialists, from architects and urban planners to game developers and landscape designers.

Objects are synthesized by a generative-adversarial neural network (GAN), which, based on a schematic segmented map, creates realistic images by borrowing parts from a model previously trained on several million photographs. In contrast to the previously developed systems of image synthesis, the proposed method is based on the use of adaptive spatial transformation followed by transformation based on machine learning. Processing a segmented map instead of semantic markup allows you to achieve an exact match of the result and control the style.

To achieve realism, two competing neural networks are used: the generator and the discriminator (Discriminator). The generator generates images based on mixing elements of real photos, and the discriminator identifies possible deviations from real images. As a result, a feedback is formed, on the basis of which the generator begins to assemble more and more qualitative samples, until the discriminator ceases to distinguish them from the real ones.