Speech signal feature extraction process: input speech signal - framing, pre emphasis, windowing, fft- > STFT spectrum (including amplitude and phase) - square the complex number - > amplitude spectru

Speech signal feature extraction process ： Input voice signal - Framing 、 Pre emphasis 、 Add windows 、FFT->STFT Spectrum （ Include amplitude 、 phase ）- Square the complex number -> Amplitude spectrum -Mel wave filtering -> MEL spectrum - Take the logarithm -> Logarithmic Mel spectrum -DCT->FBank->MFCC

---

---

---

Reference material ：
AI lemon ： Focus on speech recognition 、 Voiceprint recognition 、 Science, technology and application related to speech synthesis