arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2005.11138eess.AScs.LGcs.SDstat.ML
收藏

TinyLSTMs:用于助听器的高效神经语音增强

TinyLSTMs: Efficient Neural Speech Enhancement for Hearing Aids

Arm(Arm公司);Bose(Bose公司)

Igor Fedorov, Marko Stamenovic, Carl Jensen, Li-Chia Yang, Ari Mandell, Yiming Gan, Matthew Mattina, Paul N. Whatmough

摘要:

现代语音增强算法通过使用大型递归神经网络(RNN)实现了显著的噪声抑制。然而,大型RNN限制了助听器硬件(HW)形状因子的实际部署,这些硬件是电池供电的,并且在具有有限存储容量和计算能力的资源受限的微控制器单元(MCU)上运行。在这项工作中,我们使用模型压缩技术来弥合这一差距。我们定义了HW对RNN施加的约束,并描述了满足这些约束的方法。尽管模型压缩技术是一个活跃的研究领域,但我们是第一个使用权重/激活的修剪和整数量化来证明其对RNN语音增强的功效的。我们还演示了状态更新跳过,这减少了计算负载。最后,我们对压缩模型进行感知评估,以验证人类评分员的音频质量。结果显示,与压缩模型的基线相比,模型大小和操作分别减少了11.9 $\times$和2.9 $\times$,在收听偏好方面没有统计学差异,仅表现出0.55 dB SDR的损失。我们的模型实现了2.39毫秒的计算延迟,以及在10毫秒的目标和351 $\times$比以前的工作更好。

英文摘要:

Modern speech enhancement algorithms achieve remarkable noise suppression by means of large recurrent neural networks (RNNs). However, large RNNs limit practical deployment in hearing aid hardware (HW) form-factors, which are battery powered and run on resource-constrained microcontroller units (MCUs) with limited memory capacity and compute capability. In this work, we use model compression techniques to bridge this gap. We define the constraints imposed on the RNN by the HW and describe a method to satisfy them. Although model compression techniques are an active area of research, we are the first to demonstrate their efficacy for RNN speech enhancement, using pruning and integer quantization of weights/activations. We also demonstrate state update skipping, which reduces the computational load. Finally, we conduct a perceptual evaluation of the compressed models to verify audio quality on human raters. Results show a reduction in model size and operations of 11.9$\times$ and 2.9$\times$, respectively, over the baseline for compressed models, without a statistical difference in listening preference and only exhibiting a loss of 0.55dB SDR. Our model achieves a computational latency of 2.39ms, well within the 10ms target and 351$\times$ better than previous work.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑