{"ID":23507301,"CreatedAt":"2026-09-18T02:21:44.056544415Z","UpdatedAt":"2026-09-20T18:11:56.143995915Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2609.20463","arxiv_id":"2609.20463","title":"Beyond the Stability--Plasticity Frontier in Streaming Target Speaker Extraction","abstract":"Streaming target speaker extraction must maintain a representation of whom to extract while the target may fall silent, be masked by interference, or drift acoustically away from enrollment. Existing systems typically hold this state as a stored embedding updated by hand-designed rules. Across 22 configurations, including confidence-gated and oracle-activity-gated updates, we show that this family lies on a stability-plasticity frontier: even perfect target-activity information cannot combine robustness to target absence with adaptation to enrollment-mixture mismatch. We therefore meta-train speaker-state dynamics through the closed streaming loop, exposing the updater to its own contaminated evidence. Our proposed 41k-parameter anchored fast-weights (AFW) memory moves beyond the measured heuristic frontier, gaining 3.0 dB over the best heuristic under severe mismatch while staying within 0.9 dB of static enrollment after 30 s of absence, at under 5% runtime overhead. A gated recurrent unit (GRU) control confirms that the gain is not AFW-specific, while AFW is smaller and more interpretable: under severe mismatch, its write residual grows and aligns with the target rather than the interferer. Code is publicly available at https://github.com/ym2976/anchor-fast-weight.","short_abstract":"Streaming target speaker extraction must maintain a representation of whom to extract while the target may fall silent, be masked by interference, or drift acoustically away from enrollment. Existing systems typically hold this state as a stored embedding updated by hand-designed rules. Across 22 configurations, includ...","url_abs":"https://arxiv.org/abs/2609.20463","url_pdf":"https://arxiv.org/pdf/2609.20463v1","authors":"[\"Yuesheng Ma\",\"Linyang He\",\"Nima Mesgarani\"]","published":"2026-09-17T14:24:58Z","proceeding":"eess.AS","tasks":"[\"eess.AS\"]","methods":"[\"Large Language Model\"]","has_code":false,"code_links":[{"ID":639812,"CreatedAt":"2026-09-18T02:21:44.056544415Z","UpdatedAt":"2026-09-18T02:21:44.056544415Z","DeletedAt":null,"paper_id":23507301,"paper_url":"https://arxiv.org/abs/2609.20463","paper_title":"Beyond the Stability--Plasticity Frontier in Streaming Target Speaker Extraction","repo_url":"https://github.com/ym2976/anchor-fast-weight","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}
