MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641096213 A) filed by Elizabeth Vinitha P V; and Dr. Pramod Pavithran on August 09, 2026, for A Computer-Implemented System And Method For Adaptive Multimodal Screening Of Handwriting Difficulty.
Inventors include Elizabeth Vinitha P V; and Dr. Pramod Pavithran.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The invention relates to computer-implemented educational and medical screening of handwriting difficulty using offline handwriting images from ordinary paper digitized by a scanner or camera, together with questionnaire evidence. Task-specific regions from copied-word writing, dictated-word writing, and paragraph writing are processed in parallel by branches using Local Ternary Pattern, Discrete Cosine Transform, wavelet features, and MobileViT-S-CBAM deep features, with branch outputs converted to probabilities on a common scale. A technical advancement is provided by development-derived branch reliability estimation and Reliability-Aware Adaptive Probability Fusion, which adapt branch weights per student using both learned reliability and branch confidence, followed by uncertainty-governed fusion with a transformed questionnaire score to generate a final composite screening score and label. The invention enables low-cost, sensor-free early screening and structured reporting of handwriting difficulty in practical school or clinic settings.
Disclaimer: Curated by HT Syndication.
