MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641073664 A) filed by Brindha S on June 13, 2026, for Ai-Based Multimodal Virtual Fashion Interaction System With Real-Time Multilingualvoice Integration.
Inventors include Brindha S; Karpaga Varshiniv; Subhashini N; Abirami P; Divya M; Swathi G; Harini P; Kiruthikraj S S; Dharshini S R; Nandha K; Pranisa R; Pragadeeshwaran S; Keerthika M; Mukeshn P; Shalva S; and Nishit V P.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: The present invention discloses a unified multimodal artificial intelligence system for personalized virtual fashion interaction, integrating real-time visual analysis and speech-based communication within a single cohesive and adaptive processing architecture. The system is configured to receive multimodal user input comprising image data and speech signals, wherein the image data is processed to extract visual attributes including skin tone, body structure, pose estimation, and appearance- based features, and the speech input is processed through speech recognition, language identification, neural machine translation, and semantic analysis to derive linguistic intent, user preferences, contextual parameters, and interaction commands. A multimodal fusion engine is configured to combine the extracted visual attributes and linguistic information into a unified user representation, wherein the linguistic intent, contextual parameters, and user preferences derived from the speech input dynamically and continuously influence garment selection, fitting parameters, and visualization outcomes in real time. The fusion process enables cross-modal dependency modeling, ensuring that visual compatibility and user intent are jointly optimized within a single decision framework. Based on this fused representation, an intelligent decision engine generates optimized garment recommendations by considering visual compatibility, user-defined preferences, contextual requirements such as occasion and environment, and evolving fashion trends. The system further incorporates a virtual visualization module configured to perform realistic garment fitting on a three-dimensional avatar using geometric alignment, scaling, deformation, and adaptive warping techniques, thereby enabling interactive two-dimensional and three dimensional visualization from multiple viewpoints with improved realism and accuracy. In parallel, a multilingual voice generation module converts the generated recommendations into natural-sounding speech using neural text-to-speech synthesis, with optional voice cloning to preserve speaker-specific vocal characteristics and enhance personalization. The speech input further acts as a real-time control interface, continuously driving adaptive modification of garment selection, visual rendering parameters, and recommendation outputs within the unified processing pipeline, thereby enabling seamless voice-driven interaction. The system additionally incorporates content analytics and learning mechanisms, including sentiment analysis, keyword extraction, behavioral pattern recognition, and engagement prediction, to refine recommendation accuracy and enable adaptive personalization over time. The architecture is implemented using a scalable, cloud- integrated framework supporting asynchronous and distributed processing for efficient resource utilization and real-time responsiveness. The invention is characterized by the tight and interdependent integration of visual intelligence and speech intelligence within a single unified pipeline, wherein both modalities are inseparably fused to produce synchronized visual and auditory outputs. This approach eliminates the limitations of conventional standalone systems, enhances multilingual accessibility, enables natural human-computer interaction, and significantly reduces uncertainty in digital garment selection. The proposed system thereby represents a technological advancement in multimodal AI-driven interaction, providing an intelligent, adaptive, and immersive solution for next-generation virtual fashion and interactive e- commerce platforms.
Disclaimer: Curated by HT Syndication.
