Self-tuning Spectral Clustering For Speaker Diarization

Abstract

Spectral clustering has proven effective in grouping speech representations for speaker diarization tasks, although post-processing the affinity matrix remains difficult due to the need for careful tuning before constructing the Laplacian. In this study, we present a novel pruning algorithm to create a sparse affinity matrix called spectral clustering on p-neighborhood retained affinity matrix (SC-pNA). Our method improves on node-specific fixed neighbor selection by allowing a variable number of neighbors, eliminating the need for external tuning data as the pruning parameters are derived directly from the affinity matrix. SC-pNA does so by identifying two clusters in every row of the initial affinity matrix, and retains only the top p % similarity scores from the cluster containing larger similarities. Spectral clustering is performed subsequently, with the number of clusters determined as the maximum eigengap. Experimental results on the challenging DIHARD-III dataset highlight the

Stats

• citations1
• S2 citations—
• github stars5
• HF likes0
• heat score3.81
• arxiv keyraghav2024self

Related papers

• Auto-tuning Spectral Clustering For Speaker Diarization Using Normalized Maximum Eigengap (2020)14.58
• Assessing The Robustness Of Spectral Clustering For Deep Speaker Diarization (2024)3.58
• Spectral Clustering-aware Learning Of Embeddings For Speaker Diarisation (2022)2.26
• Multi-class Spectral Clustering With Overlaps For Speaker Diarization (2020)10.35
• LSTM Based Similarity Measurement With Spectral Clustering For Speaker Diarization (2019)13.79
• Enhancements For Audio-only Diarization Systems (2019)0.00
• Deep Self-supervised Hierarchical Clustering For Speaker Diarization (2020)5.24
• Self-supervised Representation Learning With Path Integral Clustering For Speaker Diarization (2021)8.35