SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING

Description

G10L encompasses techniques for analyzing, synthesizing, and processing speech and audio signals, including speech recognition systems that identify and interpret spoken content, as well as voice processing methods such as speaker identification and emotion detection. This subclass also covers audio and speech coding/decoding technologies that compress and decompress voice data for transmission or storage, enabling efficient communication and multimedia applications. The classification includes acoustic feature extraction, signal processing algorithms, and computational methods applied to speech and voice signals, but excludes general audio signal processing unrelated to speech (covered under G10H and related classes) and non-acoustic speech analysis such as written text processing.

Filing statistics

CPC subclass G10L recorded 175,043 patent applications worldwide between 2013 and 2023. Annual filings grew from 7,834 in 2013 to 17,369 in 2023 (+122%), peaking at 21,895 applications in 2020. The most active applicants in class G10 are SAMSUNG ELECTRONICS COMPANY (6,277 applications), GOOGLE (5,429 applications) and FRAUNHOFER (4,984 applications).

Source: EPO PATSTAT, worldwide filings