Cross-Subject Generalization in Decoding Perceived Speech from Non-Invasive Brain Recordings
从非侵入式脑记录中解码感知语音的跨被试泛化
机构 * School of Intelligence Science and Technology, Peking University(北京大学智能科学与技术学院) ; Academy for Advanced Interdisciplinary Studies, Peking University(北京大学前沿交叉学科研究院) ; National Key Laboratory of General Artificial Intelligence(通用人工智能国家重点实验室)
专题命中 BCI数据与评测 :EEG(abstract,abstract_cn)
AI总结 该研究提出CPSD框架,结合对比学习、PESA模块等,在三类感知语音数据集上跨被试解码性能优于基线,提升了Top-10准确率。
Comments Submitted to IEEE Transactions on Audio, Speech, and Language Processing (TASLP)