MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202641007161 A) filed by Dr. Anil Pandurang Gaikwad; and Dr. Balajee Maram, Warangal, Telangana, on Jan. 24, for 'an ai-driven multimodal sensor fusion system for accurate human activity classification.'
Inventor(s) include Dr. Anil Pandurang Gaikwad; and Dr. Balajee Maram.
The application for the patent was published on Feb. 6, under issue no. 06/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses an artificial intelligence-based multimodal sensor fusion system for accurate human activity classification. The system collects data from multiple heterogeneous sensors, including wearable, ambient, and device-embedded sensors. The sensor data is preprocessed through synchronization, normalization, and noise reduction. Modality-specific artificial intelligence models extract representative features from each sensor stream. The extracted features are combined using an AI-driven fusion mechanism that learns temporal and cross-modal relationships. The fusion process dynamically adapts to sensor reliability and contextual relevance. A classification module analyzes the fused representation to identify one or more human activities. The system supports real-time operation and scalable deployment. Adaptation and personalization may be performed to improve performance over time. The invention is applicable to healthcare, smart environments, fitness monitoring, and human-computer interaction."
Disclaimer: Curated by HT Syndication.
