nwnlp / cnn-asrLinks
a simple asr system
☆14Updated 7 years ago
Alternatives and similar repositories for cnn-asr
Users that are interested in cnn-asr are comparing it to the libraries listed below
Sorting:
- 2018年7⽉30⽇-8⽉13⽇持续2周的好未来AI训练营中语⾳情感识别营的项目报告☆33Updated 6 years ago
- 基于卷积神经网络的语音识别声学模型的研究☆178Updated 6 years ago
- 用于存储NLP常用模型☆146Updated 5 years ago
- TestB榜第10的方案,bleu32.1☆64Updated 5 years ago
- Seq2Seq Speech Recognition with Transformer on Mandarin Chinese☆117Updated 5 years ago
- 个人博客 (https://fancyerii.github.io/) 文章的代码☆185Updated last year
- 2018年7⽉30⽇-8⽉13⽇持续2周的AI训练营中语⾳情感识别营的项目报告。☆96Updated 7 years ago
- 这是一个基于全卷积神经网络的语音识别系统☆79Updated 6 years ago
- soft_mask_bert model for Chinese Spelling Correction in keras☆23Updated 4 years ago
- 用于机器学习的语音特征提取,包含FBank和MFCC等,原理讲解和step by step的实现☆53Updated 6 years ago
- 基于 PyTorch 范例 实现中文语言模型。☆14Updated 7 years ago
- Use bert to predict punctuation on IWSLT2012 and The People's Daily 2014☆65Updated 5 years ago
- python | 高效使用统计语言模型kenlm:新词发现、分词、智能纠错等☆168Updated 6 years ago
- Tensorflow version of DFSMN☆49Updated 7 years ago
- ASR中文语音识别☆37Updated 6 years ago
- Use ctc to do chinese speech recognition by keras / 通过keras和ctc实现中文语音识别☆43Updated 7 years ago
- 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。☆70Updated 6 years ago
- End-to-end translation of Chinese phonetics to characters using bi-directional RNN (LSTM/GRU)☆29Updated 6 years ago
- keras+tensorflow实现的各种神经网络☆86Updated 6 years ago
- 用多层BLSTM模型同时进行中文分词和标点符号预测☆18Updated 11 months ago
- 基于DNN神经网络的简单语音唤醒☆12Updated 6 years ago
- Automatic Speech Recognition with TensorFlow(CNN+BLSTM+CTC)☆12Updated 7 years ago
- End-to-end speech recognition on AISHELL dataset.☆33Updated 3 years ago
- transformer for ASR-systerm (via tensorflow2.0)☆114Updated 6 years ago
- machine learning algorithms and implementations☆116Updated 7 years ago
- keras+bi-lstm+crf,中文命名实体识别☆17Updated 7 years ago
- 基于深度学习的普通话语音识别☆17Updated 6 years ago
- 未来杯 语音赛道说话人识别的baseline☆49Updated 6 years ago
- ☆39Updated 4 years ago
- 基于dVector的说话人识别keras☆90Updated 4 years ago