TAAFT
Free mode
100% free
Freemium
Free Trial
Create tool

Mel-Frequency Cepstral Coefficients (MFCCs)

[mɛl ˈfriːkwənsi ˈkɛpstrəl kəʊˈɛfɪʃənts]
AI Infrastructure
Last updated: December 9, 2024

Definition

Feature extraction technique that represents the short-term power spectrum of sound.

Detailed Explanation

Mathematical representation that mimics human auditory perception, created by taking the cosine transform of the mel-scaled log power spectrum. Widely used in speech and audio processing for feature extraction.

Use Cases

Speech recognition, speaker identification, music genre classification, audio fingerprinting

Related Terms