5 papers
Exploiting Noise Inseparability for Weakly-Supervised Discriminative Speech Denoising Using Noisy Targets
Matthew Maciejewski, Samuele Cornell
Speech denoising is an often necessary step not only for human listening, but also for downstream processing by systems lacking robustness to noisy, real-world acoustic conditions.…
Single-Microphone Audio Point Source Discriminative Localization From Reverberation Late Tail Estimation
Matthew Maciejewski
Location information can be a valuable signal for audio segmentation tasks, especially as a complement to methods focusing on the content or qualities of the sources. Though audio…
Ring Mixing with Auxiliary Signal-to-Consistency-Error Ratio Loss for Unsupervised Denoising in Speech Separation
Matthew Maciejewski, Samuele Cornell
Noisy speech separation systems are typically trained on fully-synthetic mixtures, limiting generalization to real-world scenarios. Though training on mixtures of in-domain (thus o…
End-to-End Diarization utilizing Attractor Deep Clustering
David Palzer, Matthew Maciejewski, Eric Fosler-Lussier
Speaker diarization remains challenging due to the need for structured speaker representations, efficient modeling, and robustness to varying conditions. We propose a performant, c…
Improving Neural Diarization through Speaker Attribute Attractors and Local Dependency Modeling
David Palzer, Matthew Maciejewski, Eric Fosler-Lussier
In recent years, end-to-end approaches have made notable progress in addressing the challenge of speaker diarization, which involves segmenting and identifying speakers in multi-ta…