MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202621059838 A) filed by Amit Kumar Saxena; and Anupama Saxena on May 11, 2026, for System For Real-Time Multimodal Data Processing Using Graph-Based Imputation And Adaptive Fusion And Method Thereof.
Inventors include Rajeshwar Prasad; Suman Laha; Abhishek Dubey; Utpal Roy; Damodar Patel; Prashant Vaishnav; Rahul Kumar; and Aradhy Tiwari.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: The present disclosure relates to a system (100) for real-time multimodal data processing comprising a computing device (102) having a processor (104) and memory (106) executing functional modules (108). A data acquisition module (110) receives text, audio, and video data. A preprocessing module (112) cleans and temporally aligns the data. A graph construction module (114) builds a dependency graph with cross-modal correlations, adaptively updated by a graph updating module (116). An imputation module (118) applies graph-based filtering to missing values. An imputation confidence module (120) generates reliability values for imputed data. A feature extraction module (122) processes text, audio, and video branches (130, 134, 136) independently. A reliability scoring module (124) assigns confidence-based weights. A fusion module (126) generates a fused representation weighted by imputation-derived reliability scores. An output generation module (128) produces an assessment output indicative of a predicted condition of the subject.
Disclaimer: Curated by HT Syndication.
