Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces SR-CorrNet, an asymmetric encoder-decoder architecture for speech separation that employs a separation-reconstruction (SepRe) strategy within a time-frequency (TF) dual-path framework. The encoder performs initial separation, while a weight-shared decoder refines speaker-discriminative features through cross-speaker interaction. The method formulates speech separation as a correlation-to-filter problem, using spatio-spectro-temporal correlations to estimate deep filters for signal recovery, and incorporates an attractor-based dynamic split module.
Overlapping speech separation gets a boost from SR-CorrNet, which disentangles speakers earlier using a novel separation-reconstruction strategy and correlation-based filtering, leading to state-of-the-art results in noisy and reverberant environments.
Speech separation in realistic acoustic environments remains challenging because overlapping speakers, background noise, and reverberation must be resolved simultaneously. Although recent time-frequency (TF) domain models have shown strong performance, most still rely on late-split architectures, where speaker disentanglement is deferred to the final stage, creating an information bottleneck and weakening discriminability under adverse conditions. To address this issue, we propose SR-CorrNet, an asymmetric encoder-decoder framework that introduces the separation-reconstruction (SepRe) strategy into a TF dual-path backbone. The encoder performs coarse separation from mixture observations, while the weight-shared decoder progressively reconstructs speaker-discriminative features with cross-speaker interaction, enabling stage-wise refinement. To complement this architecture, we formulate speech separation as a structured correlation-to-filter problem: spatio-spectro-temporal correlations computed from the observations are used as input features, and the corresponding deep filters are estimated to recover target signals. We further incorporate an attractor-based dynamic split module to adapt the number of output streams to the actual speaker configuration. Experimental results on WSJ0-2/3/4/5Mix, WHAMR!, and LibriCSS demonstrate consistent improvements across anechoic, noisy-reverberant, and real-recorded conditions in both single- and multi-channel settings, highlighting the effectiveness of TF-domain SepRe with correlation-based filter estimation for speech separation.