AEC_DeepModel

基于深度学习的声学回声消除基线代码

具体解析见我的博客：https://www.cnblogs.com/LXP-Never/p/14779360.html

数据准备

按照以下文件结构，放好语音，我直接使用的是AEC-Challenge 数据集中的合成数据集

└─Synthetic
    ├─TEST
    │  ├─echo_signal
    │  ├─farend_speech
    │  ├─nearend_mic_signal
    │  └─nearend_speech
    ├─TRAIN
    │  ├─echo_signal
    │  ├─farend_speech
    │  ├─nearend_mic_signal
    │  └─nearend_speech
    └─VAL
        ├─echo_signal
        ├─farend_speech
        ├─nearend_mic_signal
        └─nearend_speech

数据处理脚本为 data_preparation.py

如果想要自己生成回声的话建议使用 RIR-Generator 方法，毕竟很多论文中使用的也是这个方法

运行

python train.py

具体的命令行解析参数见train.py脚本

估计近端语音

python test/model_test.py

点赞，关注，不迷路

我以后还会开源更有价值的内容

参考

语音数据增强及python实现

AEC_DeepModel - Deep learning based acoustic echo cancellation baseline code

Related tags

Overview

AEC_DeepModel

数据准备

运行

估计近端语音

参考

Owner

凌逆战

Grapheme-to-phoneme (G2P) conversion is the process of generating pronunciation for words based on their written form.

translate using your voice

Tevatron is a simple and efficient toolkit for training and running dense retrievers with deep language models.

This repository contains the code for running the character-level Sandwich Transformers from our ACL 2020 paper on Improving Transformer Models by Reordering their Sublayers.

A website which allows you to play with the GPT-2 transformer

Korea Spell Checker

A Non-Autoregressive Transformer based TTS, supporting a family of SOTA transformers with supervised and unsupervised duration modelings. This project grows with the research community, aiming to achieve the ultimate TTS.

A combination of autoregressors and autoencoders using XLNet for sentiment analysis

运小筹公众号是致力于分享运筹优化(LP、MIP、NLP、随机规划、鲁棒优化)、凸优化、强化学习等研究领域的内容以及涉及到的算法的代码实现。

Natural Language Processing Specialization

Google's Meena transformer chatbot implementation

Code for the paper PermuteFormer

This converter will create the exact measure for your cappuccino recipe from the grandiose Rafaella Ballerini!

Sorce code and datasets for "K-BERT: Enabling Language Representation with Knowledge Graph",

Prompt-learning is the latest paradigm to adapt pre-trained language models (PLMs) to downstream NLP tasks

AI and Machine Learning workflows on Anthos Bare Metal.

Protein Language Model

Jarvis is a simple Chatbot with a GUI capable of chatting and retrieving information and daily news from the internet for it's user.

The code for the Subformer, from the EMNLP 2021 Findings paper: "Subformer: Exploring Weight Sharing for Parameter Efficiency in Generative Transformers", by Machel Reid, Edison Marrese-Taylor, and Yutaka Matsuo

Scikit-learn style model finetuning for NLP