ClearVoice: Speech Enhancement

ClearVoice (Github Repo) is AI-powered and extracts clear speech from background noise for enhanced speech quality. It supports both 16 kHz and 48 kHz audio outputs. To try it, simply upload your audio, or click one of the examples.

FRCRN: Boosting Feature Representation Using Frequency Recurrence for Monaural Speech Enhancement (ICASSP 2022)

MossFormer2: Combining Transformer and RNN-Free Recurrent Network for Enhanced Time-Domain Monaural Speech Separation (ICASSP 2024)

ClearVoice: Speech Separation

ClearVoice (Github Repo) is powered by AI and separates individual speech from mixed audio. It supports 16 kHz and two output streams. To try it, simply upload your audio, or click one of the examples.

MossFormer: Pushing the Performance Limit of Monaural Speech Separation using Gated Single-Head Transformer with Convolution-Augmented Joint Self-Attentions (ICASSP 2023)

MossFormer2: Combining Transformer and RNN-Free Recurrent Network for Enhanced Time-Domain Monaural Speech Separation (ICASSP 2024)

ClearVoice: Speech Super Resolution

ClearVoice (Github Repo) is AI-powered and transform low-resolution audio (effective sampling rate ≥ 16 kHz) into crystal-clear, high-resolution audio at 48 kHz. It supports most of audio types. To try it, simply upload your audio, or click one of the examples.

FRCRN: Boosting Feature Representation Using Frequency Recurrence for Monaural Speech Enhancement (ICASSP 2022)

MossFormer2: Combining Transformer and RNN-Free Recurrent Network for Enhanced Time-Domain Monaural Speech Separation (ICASSP 2024)

HiFi-SR: A Unified Generative Transformer-Convolutional Adversarial Network for High-Fidelity Speech Super-Resolution (ICASSP 2025)

ClearVoice: Audio-Visual Speaker Extraction

ClearVoice (Github Repo) is AI-powered and extracts each speaker's voice from a multi-speaker video using facial recognition. To try it, simply upload your video, or click one of the examples.