MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641104332 A) filed by Dr. P. Nagabhushanam; Dr. Pulipati Nagaraju; Dr. Bhanu Pratap Reddy B; and Dr. M Shiva Rama Krishna on August 30, 2026, for Facial Emotion Recognition In Unconstrained Videos Using Multi-Scale Feature Aggregation.
Inventors include Dr. P. Nagabhushanam; Dr. Pulipati Nagaraju; Dr. Bhanu Pratap Reddy B; and Dr. M Shiva Rama Krishna.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: The present invention relates to a facial emotion recognition system for unconstrained video environments using multi-scale feature aggregation and deep learning- based temporal modeling techniques. The system processes video input containing human facial expressions captured under varying real-world conditions including changes in pose, lighting, occlusion, and background complexity. A Multi-Scale Convolutional Neural Network extracts hierarchical facial features at different resolutions to capture both fine-grained and global expression details. These spatial features are further processed using a sequence-to-sequence model that analyzes temporal dependencies across video frames to ensure consistent emotion interpretation over time. The system classifies emotions such as happiness, sadness, anger, fear, surprise, and neutrality using a probabilistic classification layer. The invention enhances robustness, accuracy, and generalization in real-world applications. It supports real-time processing and is applicable in surveillance, healthcare monitoring, automotive driver assistance, entertainment analytics, and human-computer interaction systems, enabling intelligent and adaptive emotion-aware computing across diverse domains.
Disclaimer: Curated by HT Syndication.
