Naist Contribution To Zerospeech 2017 (Track 1) - Preliminary

Name: Naist Contribution To Zerospeech 2017 (Track 1) - Preliminary
Creator: NAIST
Published: 2017-06-23T00:00:00+08:00

Zenodo2017-06-23 更新2026-04-07 收录

零资源语音处理

语音单元发现

数据链接：

https://zenodo.org/record/816326 数据链接链接失效反馈

官方服务：

资源简介：

This is the PRELIMINARY submission of NAIST for track 1 of the zero resource speech challenge 2017 (ZeroSpeech2017).

提供机构：

NAIST

创建时间：

2017-06-23

Composite Embedding Systems Based on DNN-HMM and Attention End-To-End for ZeroSpeech2017 track1 (1)

零资源语音处理

无监督声学单元发现

Deep neural networks (DNNs) were trained for posterior and bottleneck features using Japanese and other language speech data. We explore various DNN types, their combinations, and dimension reduction

NIAID Data Ecosystem70

Pairwise Learning Using Unsupervised Bottleneck Features For Zero-Resource Speech Challenge 2017 (System 2)

零资源语音处理

无监督特征学习

The system is for track1 alone. We trained an antoencoder using unsupervised bottleneck features with word-pair information from unsupervised term detection (UTD) only on the give ENGLISH corpus. The

Zenodo2017-06-15 更新40

Pairwise Learning Using Unsupervised Bottleneck Features For Zero-Resource Speech Challenge 2017 (System 2)

零资源语音处理

无监督词项发现

The system is for track1 alone. We trained an antoencoder using unsupervised bottleneck features with word-pair information from unsupervised term detection (UTD) only on the give ENGLISH corpus. The

Zenodo2020-09-18 更新20

Pairwise Learning using Unsupervised Bottleneck Features for Zero-Resource Speech Challenge 2017 (System 2)

零资源语音处理

无监督特征学习

The system is for track1 alone. We trained an antoencoder using unsupervised bottleneck features with word-pair information from unsupervised term detection (UTD) only on the give ENGLISH corpus. The

NIAID Data Ecosystem70

Japanese Fake News Dataset

虚假新闻检测

多语言文本分类

Japanese Fake News Dataset Read our project page for more details: https://hkefka385.github.io/dataset/fakenews-japanese/

Zenodo2022-01-09 更新30