Intellectual Property India Publishes Patent Application for 'Vision-Language Model System For Automated Recipe Generation From Ingredient Images' Filed by Vellore Institute Of Technology

MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541122937 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 5, 2025, for 'vision-language model system for automated recipe generation from ingredient images.'

Inventor(s) include Dr. V. Sivakumar; Mr. Kshitiz Pandey; and Ms. Kashish Tyagi.

The application for the patent was published on Jan. 2, under issue no. 01/2026.

According to the abstract released by the Intellectual Property India: "The present disclosure provides a vision-language model system for automated recipe generation from ingredient images. The system includes a vision processing module configured to receive an image containing food ingredients and to detect and classify the ingredients, a recipe generation module having a language model configured to generate recipes based on the detected ingredients, and integration between the vision processing module and the recipe generation module configured to align visual ingredient detection with textual recipe generation. The vision processing module includes a hierarchical Vision Transformer architecture having a first Vision Transformer model for macro-categorization of ingredients and a second Vision Transformer model for fine-grained ingredient classification. The system further includes a conversational AI interface configured to receive user queries and recipe modification requests and to enable real-time interactive refinement of generated recipes."

Disclaimer: Curated by HT Syndication.