MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202611081800 A) filed by Credenta Technologies Private Limited on July 02, 2026, for Confidence-Adaptive Authentication Thresholding System For Mobile Physiological Authentication Using Piecewise-Linear Threshold Adaptation With Measurement Confidence Scalar Minimum Confidence Floor Abstention And Illness Compensation Factor.
Inventors include Kumar, Ravit; and Mann, Sakshi.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: A computer-implemented confidence-adaptive authentication thresholding system (1100), operatively coupled to a physiological sensor subsystem, receives a multi- dimensional confidence vector (1102) derived from physiological signal measurements and comprising a signal quality indicator (1104), a personalisation confidence indicator (1106), an illness-state indicator (1108), and a physiological-burden indicator (1110). A confidence aggregation module (1112) computes a scalar confidence score (1114) by a configurable aggregation method. A threshold adaptation module (1116) computes a dynamic acceptance threshold (1118) as a confidence-responsive function of the score, increasing the acceptance region above a first configurable boundary such that a lower authentication score suffices, and decreasing it below a second configurable boundary such that a higher authentication score is required. An illness compensation module (1120) applies a configurable multiplicative factor when illness is detected, widening the acceptance region for genuine unwell users. A comparison and gating module (1124) compares an authentication score (1126) against the compensated threshold, emits an accept signal (1128) where the score suffices, and routes execution to a secondary authentication pathway by emitting a structured abstention signal (1130) where the confidence score falls below a configurable minimum floor. Upon the confidence score falling below the said floor, the comparison and gating module (1124) transmits a sensor reconfiguration command that alters at least one physical acquisition parameter of the physiological sensor subsystem, being one or more of an illumination intensity, an exposure setting, a frame-rate setting, and a region-of-interest setting, to re-acquire a higher-integrity signal, the decision stage and the sensor subsystem thereby cooperating in a closed loop. The closed-loop system physically reconfigures the sensor hardware, thereby distinguishing the present disclosure from conventional feedback control systems that merely adjust processing parameters. In a preferred embodiment the boundaries, slopes, floor, and compensation factor adopt specific configurable values and the function is piecewise-linear; the present disclosure in its broadest aspect is not limited to any such value, aggregation method, or function form. Optional alternative aggregation methods and adaptation curves (sigmoid, exponential, reinforcement- learning-learned, Bayesian-regression-learned) are disclosed, and the illness compensation module operates in combination with the confidence aggregation, adaptation, and floor-abstention modules as a single inventive concept. The system optionally integrates with a five-component SQI, an SQI-adaptive Kalman filter (R = R_base / SQI), a five-state HMM, a multi-tier trust architecture, and structured abstention codes, with events recorded in an encrypted audit chain. FIG. 11.
Disclaimer: Curated by HT Syndication.
