AIbase
EN
Home
Convtasnet Libri2Mix Sepnoisy 16k
C

Convtasnet Libri2Mix Sepnoisy 16k

Developed by JorisCos
ConvTasNet model trained on the Asteroid framework for noisy speech separation tasks, trained on the Libri2Mix dataset.
Sound Separation
PyTorch
#Noisy speech separation#16kHz audio processing#ConvTasNet architecture
Downloads 8,407
Release Time : 3/2/2022

Model Overview

This model adopts the ConvTasNet architecture, specifically designed to separate clear speech signals from noisy mixed audio, suitable for speech enhancement and separation tasks.

Model Features

Efficient speech separation
Capable of effectively separating speech signals in noisy environments, improving speech clarity.
Optimized ConvTasNet architecture
Utilizes 8 blocks and 3 repetitions of mask network structure with skip connections to optimize separation performance.
High-quality training data
Trained on Libri2Mix and WSJ0 Hipster Ambient Mixtures datasets to ensure model generalization capability.

Model Capabilities

Noisy speech separation
Speech enhancement
Multi-speaker separation

Use Cases

Speech processing
Speech enhancement
Extracts clear speech signals in noisy environments, suitable for speech recognition preprocessing.
SI-SDR improvement of 12.55dB, STOI improvement of 0.224
Meeting transcription
Separates different speaker voices in meeting recordings to improve transcription accuracy.
SIR improvement of 24.37dB
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase