MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202621061157 A) filed by Lakshmi Narain College Of Technology Excellence on May 14, 2026, for System And Method For Hybrid Adversarial Attack Detection In Machine Learning Inference Pipelines.
Inventors include Abhishek Bhargava; and Tanisha Singh.
The application for the patent was published on July 10, 2026, under issue no. 28/2026.
Abstract: ABSTRACT System and Method for Hybrid Adversarial Attack Detection in Machine Learning Inference Pipelines The present invention relates to a method and apparatus for the hybrid detection of adversarial attacks in machine learning inference pipelines. The present method comprising analyzing incoming data based on reconstruction deviations, similarity computations, frequency domain analysis, and statistical density estimation through an input perturbation detection module, comparing the internal neural activations of a model to a baseline profile using divergence analysis through a model integrity monitoring module, adapting detection parameters for both modules using an adaptive learning module, computing a compound score representing threat levels through a decision fusion module, and classifying inputs into normal, suspicious, and adversarial categories. The method allows for the operation of the apparatus as a model-agnostic middleware component, providing real-time deployment without affecting existing machine learning models. The apparatus guarantees adversarial detection capabilities and minimizes false detections while implementing adaptive actions.
Disclaimer: Curated by HT Syndication.
