MPSA-Conformer-CTC/Attention:一种高精度,低复杂度的端到端方法,用于西藏语音识别
Changlin Wu1,2, Huihui Sun1,2, Kaifeng Huang1,2
1School of Mechanical and Electrical Engineering, Huainan Normal University, Huainan 232001, China.
Sensors (Basel, Switzerland)
|November 9, 2024
概括
这项研究使用新型的Conformer-CTC/Attention模型与可能的稀疏注意力和MaxEnt优化来增强西藏语音识别. 该MPSA-Conformer-CTC/Attention模型显著降低了文字错误率和计算需求.
科学领域:
- 人工智能的人工智能
- 语音处理 语音处理
- 计算语言学 计算语言学
背景情况:
- 西藏语音识别面临着精度低,计算成本高的挑战.
- 端到端深度学习网络提供潜在的解决方案,但需要有效的架构和解码策略.
研究的目的:
- 开发一个改进的端到端的西藏语音识别模型.
- 通过建立机制的新型集成来提高准确性和降低计算需求.
主要方法:
- 使用Conformer架构进行特征提取.
- 综合连接式时间分类 (CTC) 和注意机制,用于联合解码.
- 引入了概率论的稀疏注意力,以解决趋同问题.
- 实现了CTC的最大优化,以提高训练稳定性.
主要成果:
- 拟议的MPSA-Conformer-CTC/Attention模型在西藏数据集上实现了显著的文字错误率降低,分别为10.68%和9.57%.
- 与基线模型相比,证明减少了记忆消耗和训练时间.
- 展示了改进的概括能力和整体准确性.
结论:
- 该MPSA-Conformer-CTC/Attention模型代表了西藏语音识别的重大进步.
- 结合了Conformer,CTC,Attention,Probabilistic Sparse Attention和MaxEnt优化,有效地解决了准确性和效率方面的挑战.
更多相关视频
相关概念视频
02:07Improving Translational Accuracy
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
02:07Improving Translational Accuracy
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...