Socov: Semi-orthogonal Parametric Pooling Of Covariance Matrix For Speaker Recognition

Abstract

In conventional deep speaker embedding frameworks, the pooling layer aggregates all frame-level features over time and computes their mean and standard deviation statistics as inputs to subsequent segment-level layers. Such statistics pooling strategy produces fixed-length representations from variable-length speech segments. However, this method treats different frame-level features equally and discards covariance information. In this paper, we propose the Semi-orthogonal parameter pooling of Covariance matrix (SoCov) method. The SoCov pooling computes the covariance matrix from the self-attentive frame-level features and compresses it into a vector using the semi-orthogonal parametric vectorization, which is then concatenated with the weighted standard deviation vector to form inputs to the segment-level layers. Deep embedding based on SoCov is called ``sc-vector''. The proposed sc-vector is compared to several different baselines on the SRE21 development and evaluation sets. The sc-

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyli2025socov

Related papers

• Attentive Statistics Pooling For Deep Speaker Embedding (2018)18.88
• Deep Speaker Embedding Learning With Multi-level Pooling For Text-independent Speaker Verification (2019)0.00
• Exploring A Unified Attention-based Pooling Framework For Speaker Verification (2018)6.77
• Study On The Temporal Pooling Used In Deep Neural Networks For Speaker Verification (2021)5.84
• Aca-net: Towards Lightweight Speaker Verification Using Asymmetric Cross Attention (2023)0.00
• Spatial Pyramid Encoding With Convex Length Normalization For Text-independent Speaker Verification (2019)8.82
• Speaker Sincerity Detection Based On Covariance Feature Vectors And Ensemble Methods (2019)0.00
• Recursive Attentive Pooling For Extracting Speaker Embeddings From Multi-speaker Recordings (2024)2.26