MUMBAI, India, Jan. 3 -- Intellectual Property India has published a patent application (202511102897 A) filed by Lovely Professional University, Phagwara, Punjab, on Oct. 25, 2025, for 'hybrid quantum split learning system for distributed collaborative model training.'
Inventor(s) include Dr. Namita Kaur; Sanjay Sood; Dr. Ankur Bahl; and Vikas Verma.
The application for the patent was published on Dec. 12, under issue no. 50/2025.
According to the abstract released by the Intellectual Property India: "A hybrid quantum split learning system for distributed collaborative model training comprising a classical client module 101 that performs partial neural network training using local data without sharing raw inputs, a hybrid quantum server module 102 incorporates quantum circuits for advanced processing and feature representation within a shared split learning model, a quantum layer design module 103 employs a qubit-efficient data-loading technique to minimize qubit usage and circuit depth, a communication interface module 104 facilitates secure exchange of intermediate activations between the classical client module 101 and the hybrid quantum server module 102, supporting scalability to at least 100 clients, an anti-reconstruction defense module applies noise-based perturbations, including dynamic quantum noise, to mitigate privacy leakage, a model evaluation module assesses classification accuracy, F1-score, and convergence stability across multiple datasets and clients."
Disclaimer: Curated by HT Syndication.
