Invention Title:

GENERATIVE FOOD DONENESS PREDICTION

Publication number:

US20240386075

Publication date:
Section:

Physics

Class:

G06F18/2413

Inventors:

Applicant:

Smart overview of the Invention

The patent application describes a method for controlling the cooking process of food items using a smart oven equipped with advanced image processing capabilities. The system captures an image of a raw food item and utilizes a generative model to create synthesized images showing the food cooked to various levels of doneness. Users can select their preferred level of doneness from these generated images, and the smart oven adjusts its cooking process accordingly.

Technical Field

The approach addresses the subjective nature of terms like "light," "medium," and "dark" when describing food doneness. By providing visual representations of cooked food at different stages, the system aims to standardize these subjective descriptions into precise, user-defined outcomes. The technology leverages generative models such as Generative Adversarial Networks (GANs) or Variational AutoEncoders (VAEs) to produce these synthesized images.

Functionality

The smart oven's processor is programmed to capture an initial raw image, generate multiple cooked versions using the generative model, and then cook the item based on the user's selection. Additionally, a doneness detection model monitors the current doneness level by comparing real-time images with the target doneness level. This monitoring ensures that the cooking process progresses towards the selected doneness level.

Cooking Cycle Control

Once a cooking cycle begins, the smart oven uses its heating system to cook the food until it reaches a minimum doneness likelihood threshold specified by the selected image. The cycle may conclude by deactivating the heating system, switching to a warming mode, or notifying the user that cooking is complete. This precise control helps achieve consistent cooking results tailored to user preferences.

Smart Oven Configuration

The smart oven incorporates various components such as a camera for capturing images, temperature sensors for monitoring internal conditions, and a user interface for interaction. The heating system includes multiple elements strategically placed within the oven cavity to ensure even cooking. This configuration allows for comprehensive control over the cooking environment, enhancing both functionality and user convenience.