MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641051258 A) filed by Nitte Meenakshi Institute Of Technology, Nitte; Dr. Sudhakar K; Mrs. Sushma K; Dr. Rubini P; Dr. Arun Kumar N; and Parameshachari Bidare Divakarachari, Bangalore, Karnataka, on April 22, for 'a multimodal framework for real-time affective computing.'
Inventor(s) include Nitte Meenakshi Institute Technology, Nitte; Dr. Sudhakar K; Mrs. Sushma K; Dr. Rubini P; Dr. Arun Kumar N; and Parameshachari Bidare Divakarachari.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a multimodal framework for real-time affective computing that overcomes the limitations of unimodal and batch-processing emotion-recognition systems of the prior art. The system comprises a video camera (101) and a microphone (102) which concurrently capture facial and vocal data of a subject, a frame pre-processor (103), an audio pre-processor (104), a face detector (105), a spectrogram generator (106), a visual convolutional neural network (107) and an auditory convolutional neural network (108) operating in parallel, a trainable multimodal fusion module (109) that applies context-sensitive confidence weighting to the outputs of the two networks, an emotion classifier (110) that emits a probability distribution over a predefined set of emotion classes, and an output interface (111). The framework is arranged to execute locally on a consumer computing device within a sub-second latency budget, is robust against deliberate emotional masking, acoustic noise and partial occlusion, and produces calibrated probabilistic output suitable for clinical behavioural analysis, adaptive e-learning, telemedicine and empathetic human-computer interaction."
Disclaimer: Curated by HT Syndication.
