An integrated speech toolkit

July 14, 2025 (1mo ago)

This article mainly documents an integrated speech toolkit and speech research platform that has a wide range of applications in summer research on artificial intelligence for speech separation.

Repo Url: https://github.com/speechbrain/speechbrain

DeepWiki: https://deepwiki.com/speechbrain/speechbrain

This toolkit was used in the model training of Mamba-Transnet ClearerVoice

Some speech metrics: SZDAx

Signal-to-Noise Ratio (SNR)

Perceptual Evaluation of Speech Quality (PESQ)

Short-Time Objective Intelligibility (STOI)

Deep Noise Suppression Mean Opinion Score (DNSMOS)

Scale-Invariant Signal-to-Distortion Ratio (SI-SDR)

And more quality benchmarks