.png){kind=icon}
返回
Fast Spectrogram Inversion Using Multi-Head Convolutional Neural Networks
.png){kind=icon}
OA
AI
DOI：10.1109/LSP.2018.2880284.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
We propose the multi-head convolutional neural network (MCNN) for waveform synthesis from spectrograms. Nonlinear interpolation in MCNN is employed with transposed convolution layers in parallel heads. MCNN enables significantly better utilization of modern multi-core processors than commonly used iterative algorithms like Griffin-Lim, and yields very fast (more than 300 x real time) runtime. For training of MCNN, we use a large-scale speech recognition dataset and losses defined on waveforms that are related to perceptual audio quality. We demonstrate that MCNN constitutes a very promising approach for high-quality speech synthesis, without any iterative algorithms or autoregression in computations.
Keyword:
reconstruction
deep learning
convolutional neural networks
short-time Fourier transform
spectrogram
time-frequency signal processing
speech synthesis
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
9.6
论文数:
1.1W
被引数:
1.7W
机构
暂无机构信息
引用论文
分享收藏Holocene climate variability in the North-Western Mediterranean Sea (Gulf of Lions)北西地中海（吕宋湾）全新世气候变率.png)
OAAI
IF0
OAAI分享收藏Synthesis of Well‐Defined Polystyrene Rink Amide Soluble Supports and Their Use in Peptide Synthesis
PREAI
分享收藏分享收藏 没有更多内容
没有更多内容