lebron-kun Goto Github PK

alibaba-mit-speech

Alibaba speech technology

asr-for-chinese-pipeline

Google Summer of Code 2018 Project: Automatic Speech Recognition for Speech-to-Text on Chinese

asr_syllable

基于卷积神经网络的语音识别声学模型的研究

asr_word

采用端到端方法构建声学模型，以字为建模单元，采用DCNN-CTC网络结构。

asrframe

An Automatic Speech Recognition Frame ，一个中文语音识别的完整框架， 提供了多个模型

asrt_speechrecognition

A Deep-Learning-Based Chinese Speech Recognition System 基于深度学习的中文语音识别系统，使用CNN、LSTM和CTC实现

attention-is-all-you-need-keras

Implementation of the Transformer architecture described by Vaswani et al. in "Attention Is All You Need"

attention-networks-for-classification

Hierarchical Attention Networks for Document Classification in PyTorch

automatic_speech_recognition

End-to-end Automatic Speech Recognition for Madarian and English in Tensorflow

chinese-novel-generation

尝试使用中文文本训练RNN来产生中文小说。

cs-notes-pdf

https://github.com/CyC2018/CS-Notes PDF版本离线阅读

deep-learning-keras-tensorflow

Introduction to Deep Neural Networks with Keras and Tensorflow

deepspeech

A PaddlePaddle implementation of DeepSpeech2 architecture for ASR.

deepspeech.mxnet

A MXNet implementation of Baidu's DeepSpeech architecture

eesen

The official repository of the Eesen project

faceai

一款入门级的人脸、视频、文字检测以及识别的项目.

fairseq

Facebook AI Research Sequence-to-Sequence Toolkit written in Python.

git-basic

hierarchical-attention-model

hierarchical attention model

kaldi

This is now the official location of the Kaldi project.

keras-transformer

Transformer implemented in Keras

keras_deepspeech2_speechrecognition

Do SpeechRecognition using DeepSpeech2, backend using Keras.

las-speechrecognition

Listen, Attend and Spell (LAS) framework for speech recognition (see https://arxiv.org/pdf/1508.01211.pdf).

listed-company-news-crawl-and-text-analysis

从新浪财经、每经网、金融界、**证券网、证券时报网上，爬取上市公司（个股）的历史新闻文本数据进行文本分析、提取特征集，然后利用SVM、随机森林等分类器进行训练，最后对实施抓取的新闻数据进行分类预测

masr

中文语音识别，高识别率预训练模型，支持docker快速安装 Chinese Speech Recognition; Mandarin Automatic Speech Recognition;

models

Models and examples built with TensorFlow

multi-class-text-classification-cnn-rnn

Classify Kaggle San Francisco Crime Description into 39 classes. Build the model with CNN, RNN (GRU and LSTM) and Word Embeddings on Tensorflow.

my_ch_speech_recognition

使用python进行语音识别,基于深度学习的中文语音识别系统

my_deep_project

个人深度学习项目整理

nmt

TensorFlow Neural Machine Translation Tutorial