Speech recognition method and system based on spectrogram and long and short term memory network in complex background noise environment

• Main Authors: BI HAOYU, FU ZUOTAO, HE GUANGHUI
• Format: Patent
• Language: Chinese; English
• Published: 16.08.2024
• Subjects: ACOUSTICS; MUSICAL INSTRUMENTS; PHYSICS; SPEECH ANALYSIS OR SYNTHESIS; SPEECH OR AUDIO CODING OR DECODING; SPEECH OR VOICE PROCESSING; SPEECH RECOGNITION

The invention discloses a voice recognition method and system based on a spectrogram and a long-short term memory network in a complex background noise environment. The method comprises the following steps: 1) acquiring an audio signal for training, and generating a corresponding spectrogram; 2) processing the spectrogram; 3) constructing a speech recognition model by using the spectrogram; 4) acquiring a to-be-identified audio signal, and generating a corresponding to-be-identified spectrogram; and 5) inputting the spectrogram to be recognized into the voice recognition model to obtain a voice recognition result. The system comprises a voice acquisition module, a signal processing module, a voice recognition module and a recognition result output module. According to the method, in a speech recognition task in a complex background noise environment, the recognition accuracy and the overall performance of a system can be remarkably improved, the robustness and adaptability of an algorithm are enhanced, and th