shenyunhang commited on
Commit
2724409
·
verified ·
1 Parent(s): dde75f8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +64 -3
README.md CHANGED
@@ -1,3 +1,64 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ ---
4
+
5
+
6
+ ---
7
+
8
+ Data source:
9
+ https://openslr.org/33/
10
+
11
+ ---
12
+
13
+
14
+
15
+ <div class="container">
16
+ <h2 class="slrStyle"> Aishell
17
+ </h2>
18
+ <p class="resource"> <b>Identifier:</b> SLR33 </p>
19
+ <p class="resource"> <b>Summary:</b> Mandarin data, provided by Beijing Shell Shell Technology Co.,Ltd
20
+ </p>
21
+ <p class="resource"> <b>Category:</b> Speech
22
+ </p>
23
+ <p class="resource"> <b>License:</b> Apache License v.2.0
24
+ </p>
25
+ <p class="resource"> <b>Downloads (use a mirror closer to you):</b> <br>
26
+ <a href="https://www.openslr.org/resources/33/data_aishell.tgz"> data_aishell.tgz </a> [15G] &nbsp; ( speech data and transcripts
27
+ ) &nbsp; Mirrors:
28
+ <a href="https://us.openslr.org/resources/33/data_aishell.tgz"> [US] </a> &nbsp;
29
+ <a href="https://openslr.elda.org/resources/33/data_aishell.tgz"> [EU] </a> &nbsp;
30
+ <a href="https://openslr.magicdatatech.com/resources/33/data_aishell.tgz"> [CN] </a> &nbsp;
31
+ <br><a href="https://www.openslr.org/resources/33/resource_aishell.tgz"> resource_aishell.tgz </a> [1.2M] &nbsp; ( supplementary resources, incl. lexicon, speaker info
32
+ ) &nbsp; Mirrors:
33
+ <a href="https://us.openslr.org/resources/33/resource_aishell.tgz"> [US] </a> &nbsp;
34
+ <a href="https://openslr.elda.org/resources/33/resource_aishell.tgz"> [EU] </a> &nbsp;
35
+ <a href="https://openslr.magicdatatech.com/resources/33/resource_aishell.tgz"> [CN] </a> &nbsp;
36
+ <br><br></p><p class="resource"><b>About this resource:</b></p><div id="about">Aishell is an open-source Chinese Mandarin speech corpus published by
37
+ Beijing Shell Shell Technology Co.,Ltd.
38
+ <p>
39
+ 400 people from different accent areas in China are invited to
40
+ participate in the recording, which is conducted in a quiet indoor
41
+ environment using high fidelity microphone and downsampled to 16kHz.
42
+ The manual transcription accuracy is above 95%, through professional
43
+ speech annotation and strict quality inspection. The data is free
44
+ for academic use. We hope to provide moderate amount of data for new
45
+ researchers in the field of speech recognition.
46
+ </p><p>
47
+
48
+ You can cite the data
49
+ using the following BibTeX entry:
50
+ </p><pre>
51
+ @inproceedings{aishell_2017,
52
+ title={AIShell-1: An Open-Source Mandarin Speech Corpus and A Speech Recognition Baseline},
53
+ author={Hui Bu, Jiayu Du, Xingyu Na, Bengu Wu, Hao Zheng},
54
+ booktitle={Oriental COCOSDA 2017},
55
+ pages={Submitted},
56
+ year={2017}
57
+ }
58
+ </pre>
59
+ </div><p class="resource"> <b>External URL:</b> <a href="http://www.aishelltech.com/kysjcp"> http://www.aishelltech.com/kysjcp </a> &nbsp; Full description from the company website
60
+ </p>
61
+
62
+ <div style="height:300px"> </div>
63
+
64
+ </div>