Welcome to KingLine Data Center!   Contact Phone: 0086-10-62660053   Email: marketing@speechocean.com

English

人工智能数据资源服务平台 Speechocean Homepage Help

Home > Help > How to Start

Welcome to KingLine Data Center

KingLine Data Center, operated by Speechocean, is a fundamental data resource sharing platform which is dedicated to serve the global users of institutes (scientific research institutions, universities), enterprises and individuals by providing high quality data resources for Artificial Intelligence Technology.

KingLine Data Center FQA:

How to Become a Member?

How to Gain KingLine Credits?

How to Use KingLine Credits?

What is KingLine Commercial Data Resource?

What is KingLine Academic Data Resource?

How to Make Payment?

How to Share Your Academic Data Resources with KingLine Data Center?

How to Distribute Your Data Resources on KingLine Data Center?

How to Become a Member?

  • Join KingLine Data Center is free, please click here to Register. Earn credits immediately.
  • After successful registered, you may Complete Profile further along with your data preference; you can also gain extra credits!

How to Gain KingLine Credits?

  • Gain 15 credits against successful registration.
  • Gain extra 30 credits by filling out your profile in details along with data preference. Fill out your profile now!
  • You can get 15 credits by referring a friend to register as our member. Please Email us by contact@speechocean.com with the user ID of your recommended member.
  • We welcome members to find out errors of the released databases. You can get 50 credits as a reward once the correction is accepted .
  • Give your comments to the corpus you have used and earn 10 credits against each valid comment. (Please avoid radical words in comments or interaction with other members, otherwise the admin will sort out the comments, some may be banned from registration.)
  • Data shared by you will be listed in the academic resources catalog. Credits wll be debited to your acount against every valid access by other members.  
  • Consign your high-commercial-valued data resources with us on commission base, you can get 100 credits after signed the distribution agreement with Speechocean.
  • The most effective way to gain credits - Share Your Data With Us!
                                               
                       How to Gain Credits by Sharing Data?

 Speech Corpus:

Recording Hours (R.H)- ASR Credits Per 0.1 Hours
 R.H≤0.5 Hours  1 Credit
 0.5<R.H≤1.0 Hours  2 Credits
 1.0<R.H≤2.0 Hours  3 Credits
 2.0<R.H≤4.0 Hours  4 Credits
 4.0<R.H≤8.0 Hours  5 Credits
 8.0<R.H≤16.0 Hours  6 Credits
 16.0<R.H≤32.0 Hours  7 Credits
 32.0<R.H≤64.0 Hours  8 Credits

Recording Hours (R.H)-TTS Credits Per 0.1 Hours
 R.H≤1.0 Hours  5 Credits
 0.99<R.H≤1.99 Hours  10 Credits
 2.0<R.H≤2.99 Hours  15 Credits
 3.0<R.H≤3.99 Hours  20 Credits
 4.0<R.H≤4.99 Hours  25 Credits
 5.0<R.H≤5.99 Hours  30 Credits

Note
  • In case of recording hour is less than 0.1 hours, it will be rounded up or rounded down. For example: 1.04 hours will be rounded down to “1.00 hour”, 1.05 hour will be rounded up to 1.1 hour.
  • Speech corpora need to be manually transcribed and labeled.
  • In case the corpus is special valued, you may contact us to clarify directly.
  • If the corpus meets one of the following facts, you will gain 10%-20% extra credits.
    • The recording time is more than 80 hours;
    • The recording platform is in-car;
    • Conversational speech corpus;
    • The corpus with special labeling;
    • Uncommon language corpus.

 Lexicon:

Entries (E.) Credits Per 100 Entries
E.≤ 5000 1 Credit
5000 < E.≤10000 2 Credit
10000< E.≤20000 3 Credit
E.≥20000 4 Credit

Note:  
  • For uncommon languages, or lexicon contains over 50000 entries, you may gain 10%-20% extra credits. Or, you may contact us to clarify directly.

 Text Corpus- Web Crawling Data

  • Every 5000 Sentences can exchange for 1 credit.
Note:  
  • The corpus must be processed by filtering out the sensitive words, noisy words, repeated chatting sentences.
  • If the corpus has been proofread manually or with precise labeling, you may get 10%-20% extra credits. Or, you may contact us to clarify directly.

 Text Corpus- SMS, Email, Chatting and etc. Data

  • Every 100 Sentences can exchange for 2 credits.
Note:  
  • The corpus must be processed by filtering out the sensitive words, noisy words, repeated chatting sentences.
  • If the corpus has been proofread manually or with precise labeling, you may get 10%-20% extra credits. Or, you may contact us to clarify directly.

 Text Corpus –Machine Translation:

Corpus Type Credits for 100 pairs of sentences
2 languages paralleled corpus 2 Credits
3 languages paralleled corpus 3 Credits
4 languages paralleled corpus 4 Credits
5 languages paralleled corpus 5 Credits

Note:  
  • For uncommon languages, or corpus contains over 50000 pairs of sentences, you may get 10%-20% extra credits. Or, you may contact us to clarify directly.

 Image Corpus:

  • Every 1M can exchange 1 credit.
Note:
  • In case the corpus is less than 0.1M, it will be rounded up or rounded down. For example: 1. 4M will round down to “1.0M”, 1.5M will round up to “2.0M”.
  • The image corpus must be labelled. 
  • If the corpus has been manually annotated with precise label/multi-point/multi-level,you may get 10%-20% extra credits. Or, you may contact us to clarify directly.

Video Corpus:

  • Every 1M can exchange 1 credit.
Note:
  • In case the corpus is less than 0.1M, it will be rounded up or rounded down. For example: 1. 4M will round down to “1.0M”, 1.5M will round up to “2.0M”.
  • The video corpus must be annotated. 
  • If the corpus has been manually annotated with precise label/multi-point/multi-level,you may get 10%-20% extra credits. Or, you may contact us to clarify directly.

How to Use KingLine credits?

  • Members can use KingLine credits to exchange academic data resources;
  • The credits cannot be refunded for cash;
  • The credits can be transferred to other members, please contact us by Email: marketing@speechocean.com and indicate the user ID you wish to transferred to.

What is KingLine Commercial Data Resource?

  • There are 500+ large-scale, high-quality commercial data resources now available in KingLine Data Center, covering 110+ languages and accents. These data are designed, collected and processed by Speechocean with complete intellectual property rights (IPR). These diversified corpora are suitable for research, model training in Artificial Intelligence technology, such as Speech Synthesis, Speech Recognition, Natural Language Processing, Machine Translation, Web Search, Pattern Recognition, etc., and have been highly praised by our customers around the world.
  • You can view the parameters, introduction and samples of each corpus online. If you have any further inquiries, please call +86-01-62660928 or email contact@speechocean.com. You will be assisted with professional consultation. We also provide customized data service solution to meet your specific demands. (For more information, please visit “One-stop Data Service Solution”.)

What is KingLine Academic Data Resources?

  • KingLine Data Center has a large number of academic corpora and most of them are designed and built-up by Speechocean with precise annotation. All these databases are of high-academic-value and can satisfy the academic research requirements fully.
  • Corpus shared by members has been properly verified by us in order to guarantee fine research value and practicability.
  • Academic corpora are provided to members with very competitive price. Members can also use the credits to exchange instead of paying by cash. Please click How to Gain KingLine Credits? to find out the variety ways to obtain credits.
  • We promote couples of academic corpora monthly, please click "Monthly Promotion"for details.

How to Make Payments:
  • For overseas members:
    • INTERMEDIARY BANK'S NAME:HSBC BANK USA NA  
    • SWIFT CODE:MRMDUS33 
    • ACCOUNT WITH BANK'S NAME:HSBC (China) Beijing Branch 
    • SWIFT CODE:HSBC CNSHBJG
    • ACCOUNT NUMBER: 626-219000-003 
    • NAME OF THE ACCOUNT: Beijing Haitian Ruisheng Science Technology Ltd 
  • For domestic members:
    • 开户行:招商银行清华园支行
    • 户    名:北京海天瑞声科技有限公司
    • 帐    号:110919851510701

How to share your academic data with KingLine Data Center:

  • If you have high quality academic data resources and MAKE SURE that you have the full legal right of disposal, you can share them with us. Of course, you will earn considerable credits. If your data is downloaded by other members, credits will be debited to your account too.
  • KingLine Data Center welcomes corpora for following application: Speech Synthesis, Speech Recognition, Natural Language Processing, Machine Translation, Web Search, Pattern Recognition, etc. The valid formats includes MP3,WAV,TXT, JPEG, etc. All the corpora must have detailed specification.
  • Steps to share your data:
    • Click “Data Sharing” on the homepage.
    • Please read the “Data Sharing Declarations” carefully.
    • Paste and fill up the application form in your email and send to us.  
    • Submit samples by one of the following options:
      • Email: marketing@speechocean.com;
      • Upload to Speechocean FTP;
      • Express (Click “About” to view courier address).
    • Samples and ownership will be verified by Speechocean technical team. It cannot be listed on KingLine Data Center in case of disqualified.
    • After your data released online, KingLine credits will be deposited in your account in 2 working days.  
    • In case you need to modify data your shared, please contact us by marketing@speechocean.com;

How to Distribute Your Database on KingLine Data Center?

  • If you have high-valued commercial database and the legal right of disposal, you can consign Kingline Data Center to distribute on commission basis. 
  • KingLine Data Center welcomes corpora of following application: Speech Synthesis, Speech Recognition, Natural Language Processing, Machine Translation, Web Search, Pattern Recognition, etc. The valid formats includes MP3,WAV,TXT, JPEG, etc. All the corpora must have detailed specification. 
  • Steps to distribute your data:
    • Click “Data Sharing” on the homepage.
    • Please read the “Data Distribution Declarations” carefully.
    • Paste and fill up the application form in your email and send to us. 
    • Submit samples by one of the following options:
      • Email: marketing@speechocean.com;
      • Upload to Speechocean FTP;
      • Express (Click “About” to view courier address).
    • Samples and ownership will be verified by Speechocean technical team. It cannot be listed on KingLine Data Center in case of disqualified. 
    • After signing of the “Data Distribution Agreement” with Speechocean, your database will be listed online. KingLine credits will be deposited in your account in 2 working days and can be used to exchange other academic data resources.