MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202621061047 A) filed by Pimpri Chinchwad College Of Engineering, Pune; Prakash Keshvrao Ukhalkar; Tanya Sharma; Gitanjali Shankar Yadav; and Divya Manoj Madrewar on May 13, 2026, for Accent-Aware Speech Processing System And Method For Speech Recognition.
Inventors include Prakash Keshvrao Ukhalkar; Tanya Sharma; Gitanjali Shankar Yadav; and Divya Manoj Madrewar.
The application for the patent was published on July 10, 2026, under issue no. 28/2026.
Abstract: ABSTRACT ACCENT-AWARE SPEECH PROCESSING SYSTEM AND METHOD FOR SPEECH RECOGNITION An accent- aware speech processing system (100) and method (200) for speech recognition are disclosed. The system (100) comprises an audio input interface (102), a pre-processing module (104) for conditioning an input audio signal, a feature extraction module (106) for deriving acoustic features, and an accent classification module (108) for determining a speaker accent prior to transcription. A routing controller (110) selects an optimal speech-to-text engine from a plurality of heterogeneous engines (112) based on a stored accent- performance mapping (114), and an output module (116) generates a transcription with an associated confidence score. The method (200) includes receiving (202) an audio signal, conditioning (204) the signal, extracting (206) acoustic features, determining (208) the speaker accent, selecting (210) a speech-to-text engine in real time, generating (212) a transcription with a confidence score, and computing (214) a bias metric, thereby reducing accent-dependent recognition error. FIG. 1 will be the reference figure.
Disclaimer: Curated by HT Syndication.
