MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541122202 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 4, 2025, for 'multimodal emotion recognition using scene graph fusion.'
Inventor(s) include Dr. T. Joshva Devadas; Dr. Naveenkumar Jayakumar; Aritra Saha; and Aryaman Kantharia.
The application for the patent was published on Jan. 2, under issue no. 01/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure provides a computer-implemented method (100) for multimodal emotion recognition. The method includes receiving (102) multimodal input data comprising video data, audio data, text data, and scene graph data. The method processes (104, 106, 108, 110) the video data to extract visual features, the audio data to extract audio features, the text data to extract textual features, and the scene graph data to extract contextual relationship features. The method combines (114) the extracted features from multiple modalities through fusion to generate fused multimodal features. The method generates (116) emotion predictions with accuracy, precision, recall, F-score, and AUC metrics based on the fused multimodal features. The system processes scene graph data by loading (202) a scene graph file, extracting (206) object-predicate-object triples, and outputting (218) a scene graph embedding vector of dimension 768."
Disclaimer: Curated by HT Syndication.
