Hot eventLive
防欺骗说话人验证:统一共享编码器与混合小波提示调优
1 reports1 sources3 hr ago updated
Get the story
AI overview
该研究面向防欺骗说话人验证,提出在单一冻结的 W2V-BERT 2.0 骨干网络上构建统一共享编码器,并采用深度混合小波提示调优(WPT)按任务适配分支,仅训练约 7M 参数。在 SpoofCeleb 评测中,报告结果为 0.03% CM-EER 与 0.038 min a-DCF,并称该结果超越双编码器级联方案(0.16% CM-EER、0.047 min a-DCF)。目前公开信息仅见该 arXiv 报道,尚未见独立复现或同行评议结果。
Generated from reports · updated 3 hr ago
Timeline
Follow the coverage from different angles.
Oct 8, 2026
- arXiv · Audio and Speech防欺骗说话人验证中的统一共享编码器与混合小波提示调优
该研究提出在单一冻结的 W2V-BERT 2.0 骨干网络上,用深度混合小波提示调优(WPT)按任务适配分支,仅训练约 7M 参数,在 SpoofCeleb 评测中以 0.03% CM-EER 和 0.038 min a-DCF 超越双编码器级联(0.16%、0.047)。
Heat trend
Current heat 9·Comparable peak 10(Oct 8)·Comparable change over 24 hours –
The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.