WHAM

The 'WHAM!' dataset is a benchmark that contains noisy speech mixtures used to evaluate the performance of speech separation and denoising systems.

Papers using WHAM (16)

Exploring Self-attention Mechanisms For Speech Separation2022 · 46 citesNoise-aware Speech Separation With Contrastive Learning2023 · 7 citesRing Mixing with Auxiliary Signal-to-Consistency-Error Ratio Loss for Unsupervised Denoising in Speech Separation2026A Study of the Scale Invariant Signal to Distortion Ratio in Speech Separation with Noisy References2025Dynamic Slimmable Networks for Efficient Speech Separation2025Attractor-Based Speech Separation of Multiple Utterances by Unknown Number of Speakers2025 · 1 citesListen to Extract: Onset-Prompted Target Speaker Extraction2025Audio-visual Speech Separation In Noisy Environments With A Lightweight Iterative Model2023 · 11 citesMulti-dimensional And Multi-scale Modeling For Speech Separation Optimized By Discriminative Learning2023 · 3 citesResource-Efficient Separation Transformer2022 · 11 citesSPMamba: State-space model is all you need in speech separation2024 · 8 citesMulti-Dimensional and Multi-Scale Modeling for Speech Separation Optimized by Discriminative Learning2023 · 1 citesNoise-Aware Speech Separation with Contrastive Learning2023 · 1 citesMossFormer2: Combining Transformer and RNN-Free Recurrent Network for Enhanced Time-Domain Monaural Speech Separation2023 · 1 citesUSEF-TSE: Universal Speaker Embedding Free Target Speaker Extraction2024 · 1 citesAudio-Visual Speech Separation in Noisy Environments with a Lightweight Iterative Model2023