VoxLock: Target-Speaker Extraction for Voice Agents Without an Enrollment Step
READ THE ORIGINAL SOURCEOpen original sourcehuggingface.co · Opens in a new tabVoxLock is a 27M-parameter streaming target-speaker extraction model for voice agents. It was fine-tuned from the public REAL-TSE model and works from a voice sample cut from the call itself, without an enrollment step. On Krisp's 265-recording benchmark it reaches 4.95% WER versus 23.21% unprocessed, and on 400 AMI meeting windows it cuts WER from 42.95% to 10.63%. The 8-bit model runs a live call on 77% of one CPU thread with 24 ms algorithmic latency. Model, code and per-file scores are released.
One primary source; other links are context, not independent corroboration.
- Primary sourceHugging Face Daily Papers
- Discovery linkHugging Face
The paper page is linked. Peer review and findings have not been independently checked by AINewsIQ.
Open full article view →- Original paper+45
- Fetched source page+30
- No editor link review+0
- Original non-discussion link+5
This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.
View full scoring →Read the scoring rubric →What the source says
VoxLock is a 27M-parameter streaming target-speaker extraction model for voice agents. It was fine-tuned from the public REAL-TSE model and works from a voice sample cut from the call itself, without an enrollment step. On Krisp's 265-recording benchmark it reaches 4.95% WER versus 23.21% unprocessed, and on 400 AMI meeting windows it cuts WER from 42.95% to 10.63%. The 8-bit model runs a live call on 77% of one CPU thread with 24 ms algorithmic latency. Model, code and per-file scores are released.
A READER'S EVIDENCE CHECKHow to read this signalSix questions about claims, sources, risks, and open checks
This is a signal-level guide until claim-level evidence is available. “Not recorded” means the evidence has not been assessed here; it does not mean a result is false. How we check AI news →
- 01Who is making each claim?
The claimant and whether each claim was measured or asserted are not recorded in this public brief.
- 02What was actually tested?
Not recorded per claim. Check the source for sample size, tasks, model and benchmark versions, and reproducibility.
- 03Where are the sources?
Original link: huggingface.co. Brief basis: fetched source page. Primary page date and quote locations are not established in this panel.
- 04What could mislead?
Source claims not independently checked; results are from the author's own benchmarks and comparisons, and the author notes a commercial comparison on a specific benchmark.
- 05What is at risk if wrong?
Not assessed per claim. Consider safety, privacy, cost, and adoption decisions before relying on the result.
- 06What would change the view?
No claim-specific check is recorded. An inspectable primary method or independent replication could change the assessment.
Key findings from the stored brief
- On Krisp's 265-recording benchmark, WER drops to 4.95% from 23.21% unprocessed.
- On 400 AMI meeting windows, WER falls from 42.95% to 10.63%.
- The 8-bit model runs a live call on 77% of one CPU thread with 24 ms algorithmic latency.
- Its WER is not significantly different from Krisp's best filters but is significantly lower than other systems tested.
Source claims not independently checked; results are from the author's own benchmarks and comparisons, and the author notes a commercial comparison on a specific benchmark.
This is a source-based abstract, not independent validation of the source's claims.
Evidence & sources
These links show where the signal came from and what the source page linked to. A linked page is context, not independent corroboration.
Related archive coverage
These relationships record a citation or topic overlap, not confirmation of the finding.
The paper page is linked. Peer review and findings have not been independently checked by AINewsIQ.
Timeline & updates
- Signal dated by source or discovery feed
This timestamp places the signal in the radar. It may be the discovery post date rather than the primary page's first publication date.
- Shared brief stored
AINewsIQ prepared one abstract from fetched source page.
A publication revision history is not stored for this signal. New engagement snapshots do not mean the source itself changed.
Why this may matter
Voice agents often transcribe bystanders as well as the caller. A small, CPU-friendly isolator that needs no enrollment could make cleaner speech recognition practical in live calls.
Limits and open questions
Source claims not independently checked; results are from the author's own benchmarks and comparisons, and the author notes a commercial comparison on a specific benchmark.
Organization-specific financial, security, and operational effects require the organization's own context and review.
Discussion
Comments are reviewed before publication. Discuss the evidence, not the person.
No published comments yet.
Reader discussion opens with accounts. You can read the original source and published comments now.