This document is for informational use only, it is subject to change without notice, and should not be construed as a commitment by Nexiwave Canada Inc. Nexiwave Canada Inc. assumes no responsibility or liability for any errors or inaccuracies that may appear in the informational content contained in this guide. No part of this guide may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, recording, or otherwise, without the prior written permission of Premiere Global. Please note that the content in this guide is protected under copyright law.

Please remember that any existing artwork or images that you may want to include in your project may be protected under copyright law. The unauthorized incorporation of such material into your new work could be a violation of the rights of the copyright owner. Please be sure to obtain any permission required from the copyright owner.

Any references to company names in sample templates are for demonstration purposes only and are not intended to refer to any actual organization.

Nexiwave is a service mark of Nexiwave Canada Inc. All other trademarks are the property of their respective owners.

1. Summary

nexiwave provides a Web Service based Speech Recognition platform. The platform features:

1. Advanced Speech Indexing, locate exact location of the spoken words

2. Search focused, very high robust search accuracy

3. Full text transcription capability for external indexing capability

4. Automatic speaker identification and segmentation

5. Very high speed GPU based solution

6. Multi-language support (English + French)

7. SaaS platform, private cloud available

1.1 Why Speech Recognition?

Over 70% of human communication is carried via speech. However, effective information retrieval within recorded speech communications has been very limited. Many innovative applications can be developed if there is a simple and effective way to retrieve information from these audio archives.

2. General

2.1 Function Overview

Nexiwave Speech Recognition Platform API can be roughly separated into these areas:

· Record Management: how the audio can be moved into Nexiwave system.

· Obtain Results: obtain speech-to-text results

2.2 Call Flow

The basic call flow between your application and Nexiwave system is as below:

2.3 Asynchronous Flow

Nexiwave process your audio asynchronously. Your server does not need to wait for the processing to finish. Nexiwave can notify your server when the processing is finished and result is ready to be retrieved.

This function is supported via httpNotifyBackURL parameter in the “add” call.

3. Common Web Service Info

3.1 Common Data

Nexiwave Speech Recognition web service APIs are all stateless API calls. You must provide these parameters for every request:

Parameter	Value	Size	Mandatory	Comments
authData	component
email	String	255	Yes	Your user account email address
passwd	string	255	Yes

3.2 RESTful API Support

Besides the standard SOAP interface, the same Nexiwave Web Service API can be accessed via simple REST address.

A sample RESTful GET URL:

https://api.nexiwave.com/services/SpeechIndexing/getTranscription?authData.email=XXX&authData.passwd=XXX&recordingId=1234

3.3 XML vs. JSON

The default response content is XML, which is suitable for many XML clients to parse with the WSDL.

To receive response as JSON objects, simply add this HTTP parameter to your REST URL:

HTTP Parameter	Value
response	application/json

Sample JSON URL:

https://api.nexiwave.com/services/SpeechIndexing/add?authData.email=XXX&authData.passwd=XXX&mediaURL=XXX&response=application/json

Sample JSON return:

{"addResponse":{"return":10805}}

4. Record Management Functions

These Nexiwave services provide methods to import speech content into Nexiwave system.

Three methods are provided: add a URL, post a file, passive Index-On-Demand URL.

4.1 Web Service: add

Your application invokes this web service to instruct Nexiwave to index a media file.

Simplified Request Parameters:

(Recommended minimum list of parameter. Please consult Full Parameter List)

*Parameter*	*Value*	*Size*	*Mandatory*	*Comments*
authData	component		Yes	See “Common Web Service Data” section
data	component
mediaURL	String	5000	Yes	An URL (http/https, ftp, please) Mandatory unless mediaFileData present.
externalKey	string	5000		The external key. Note: nexiwave simply records this data. No validation will be enforced to ensure the uniqueness of this key. If two or more recordings have same externalKey, the definition of finding the recording by this key is undefined.
keywords	String	5000		A list of keywords or phrases that might be spoken in the audio, separated by comma, or dots. You may also pass in time coded keywords/phrases, using the “Nexiwave Human Text Convention”. Integration Tips: you may obtain keywords/phrases · Video OCR (with time codes) · Words/phrases from an associated PowerPoint presentation. Important: This value is used by Nexiwave Speech engine.
displayTitle	String	5000		An optional field for display in nexiwave Web UI. Important: used by Nexiwave Speech engine.
notes	String	5000		Any notes Important: This value is used by Nexiwave Speech engine.
hints	component
speakerNames	String	1000		The speaker names, separated by semi-colon (;) Important: This value is used by Nexiwave Speech engine.

RESTFul Request:

https://api.nexiwave.com/services/SpeechIndexing/add?authData.email=XXX&authData.passwd=XXX&data.mediaURL=XXX&data.keywords=XXX&data.hints.speakerNames=XXX....

Response:

Parameter	Value	Mandatory	Comments
Recording
recordingId	long	Yes

Sample Response:

Sample XML Response	http://www.nexiwave.com/static/api/captured.soap.messages/add.response.xml
Sample JSON response	http://www.nexiwave.com/static/api/captured.soap.messages/add.response.json.txt

4.2 Post Media File

This method is useful if your media file cannot be exposed via a web or ftp server.

Request:

Method:	HTTP POST
HTTP Post URL	https://api.nexiwave.com/file/storage/you@your.email.com/recording/?authData.passwd=XXX
HTTP Body	The media file to be uploaded, the raw form.

Response:

Body

Long

the new recordingId

Supported Parameters in HTTP URL (or HTTP Header)

Parameters:

externalKey

keywords

displayTitle

notes

httpNotifyBackURL

storeAudio

processingPriority

targetDecodingConfigName

humanTranscription

· Please consult Full Parameter List for the definition of each parameter.

· These parameters can be sent either through HTTP URL Parameters or HTTP header.

· All these parameters are optional.

4.3 Web Service: remove

You may also delete one of your media record from our system. These actions will be taken:

· The database record, as well as associated audio, transcript and index files, for this recording will be deleted;

· Any current jobs will be deleted

Request:

Parameter	Value	Mandatory	Comments
authData	component	Yes	See “Common Web Service Data” section
recordingId	Long	Yes	The nexiwave id for this recording

RESTFul GET Request:

https://api.nexiwave.com/services/SpeechIndexing/remove?authData.email=XXX&authData.passwd=XXX&recordingId=XXX

Response:

Parameter	Value	Mandatory	Comments
Result	boolean	Yes	Whether the removal was successful.

Sample Web Service Request and Response:

Sample Response

http://www.nexiwave.com/static/api/captured.soap.messages/remove.response.xml

4.4 Web Service: listAllRecordingInfo

You may use this method to list all recordings of your account.

Request:

Parameter	Value	Mandatory	Comments
authData	component	Yes	See “Common Web Service Data” section

RESTFul GET Request:

https://api.nexiwave.com/services/SpeechIndexing/listAllRecordingInfo?authData.email=XXX&authData.passwd=XXX

Response:

Parameter	Value	Mandatory	Comments
Recording
recordingId	long	Yes
url	String
externalKey	String
uploadTime	Date
Title	String
notes	String
averageAccruacy	Float (0~1.0)		· 0: no confidence, or confidence not available. · 1: means extremely good confidence. · A value greater than 0.7 is considered good confidence.

Sample Web Service Request and Response:

Sample Response

http://www.nexiwave.com/static/api/captured.soap.messages/listAllRecordingInfo.response.xml

5. Obtain Results Functions:

5.1 Web Service: getTranscription

IMPORTANT:

Un-adapted multi-speaker machine speech transcription is exceptionally hard. Machine transcription is most likely not suitable for human viewing, due to lack of stable accuracy, punctuation and capitalization. Machine transcription is most likely not suitable for human eyes to view the complete chunk of the machine transcription.

The following conditions may cause less than ideal speech-to-text results:

Recording Device: different recording devices may choose/discard bands of frequency differently from the recording device that Nexiwave uses.
Acoustic Condition: less than ideal acoustic conditions, such as background noise, background music, will in general cause less ideal results.
Speaker Accent: very strong speaker accent can also cause model mismatch.

Machine transcription is suitable for these purposes:

Speech analytics and speech mining
Subtitle type of function: small chunks of text display along with the hint of the audio.

It is understood that Nexiwave un-adapted machine transcription may obtain generally acceptable raw text accuracy (73%) with the following conditions:

Good acoustic conditions (no or little background noise)
Speaker accent: good or fair standard accents

In general, Nexiwave Speech-to-text function may do well on these situations:

Formal conference recordings
Studio recordings
Clear classroom teaching recordings
Clear voice memo recordings

For situations where speech-to-text results are less ideal, Nexiwave recommends our “Audio Search” functions. Our proprietary Audio Search technology is not simple search in the generated text. Nexiwave Audio Search technology provides much higher and much stable accuracy than simple searching in text. It is ideal for retrieving information from media archives.

You invoke this service when you would like to obtain full text transcription of an audio.

Request:

Parameter	Value	Mandatory	Note
authData	component	Yes	See “Common Web Service Data” section
recordingId	long	Yes	The nexiwave recording ID
alternativeSize	int	No	Size of alternative words. Negative (Default): return all alternatives. Zero: only the best result is returned. Positive: upto this number of alternative will be returned

Sample RESTFul Request:

https://api.nexiwave.com/services/SpeechIndexing/getTranscription?authData.email=XXX&authData.passwd=XXX&recordingId=XXX

Response:

Parameter	Value	Mandatory	Note
*SentenceChunk*			A sentence
startInMS	Long	Yes	Start time in milliseconds
endInMS	Long	Yes	End time in milliseconds
*wordNode*	Component
startInMS	Long	Yes	Word Start time in milliseconds
endInMS	Long	Yes	Word End time in milliseconds
wordNodeItem	Component	Yes	The contextual transcription generated
Value	String		The word
confidence	Float		The confidence measure, from 0 to 1 (1 as most confident)

Sample Response:

Sample XML Response	http://www.nexiwave.com/static/api/captured.soap.messages/getTranscription.response.xml
Sample JSON response	http://www.nexiwave.com/static/api/captured.soap.messages/getTranscription.response.json.txt

6. Misc Functions:

6.1 Web Service: queryProcessingStatus

Depends on the length of your multimedia file, nexiwave indexing can take some time to finish. You may this web service to query the current status of your recording.

Request:

Parameter	Value	Size	Mandatory	Comments
authData	component		Yes	See “Common Web Service Data” section
recordingId	Long		Yes	The nexiwave recording ID

RESTFul GET Request:

https://api.nexiwave.com/services/SpeechIndexing/queryProcessingStatus?authData.email=XXX&authData.passwd=XXX&recordingId=XXX

Response:

Parameter	Value	Mandatory	Comments
RecordingProcessingStatus
Status	String	Yes	See “Table 1” below for possible values
extraInfo	Component		Note: this field may only be valid when the “status” is “DECODING_IN_PROGRESS”
percentageFinished	float	yes	0-100: percentage of finished
estimatedFinishedTimeInMS	long	yes	Estimated time to finish in milliseconds
otherStats	string	no	Other tech stats

Sample Web Service Request and Response:

Sample Response

http://www.nexiwave.com/static/api/captured.soap.messages/queryProcessingStatus.response.xml

6.1.1 Possible Processing Status

This table below lists all possible statuses

Table 1: Possible Processing Status Values

*Status Values*	*Meaning*	*Note*
INVALID_AUDIO_SOURCE	Nexiwave was unable to obtain audio from specified source
INVALID_RECORDING	The recording downloaded, but the content is not valid media content.	Please refer the support media format section for a list of supported media format.
RECORD_NOT_EXIST	The recording not exist
WRITING_RECORD	The recording is being written to DB
RECEIVING_ORIG_MEDIA	Starting to receive original media content
READY_4_AUDIO_EXTRACTION	The upload request is in queue, audio source has been obtained. Audio is ready to be extracted.
EXTRACTING_AUDIO	nexiwave is still extracting audio from the original media file
NOT_FOUND_IN_QUEUE	Audio is ready, but the recording is not in the processing queue yet	Nexiwave prioritize the processing based on three levels: · Paying customer with prioritized processing · Paying customer · Free trial customers
READY_FOR_DECODING	The recording has been scheduled to be processed and waiting in processing queue.
QUEUED_FOR_DECODING	Recording is queued for processing, but not started yet
DECODING_IN_PROGRESS	Decoding has started	extraInfo field contains the current decoding progress: · % of finished · estimated finish time · decoding speed
IN_FINAL_PROCESSING	Decoding has finished, in final verification and writing data.
FINISHED	Decoding has finished
WORD_SPOTTING_LIST_MATCHED	Some spotting words have been found in the audio (Decoding should have been nearly finished)	extraInfo contains: · number of hits (an integer)
TIME_STAMPING_DONE	Time stamping process has finished for the recording.
UNKNOWN	The status of this recording is unknown
FAILED	Nexiwave failed to process this recording

6.1.2 Typical Processing Status Sequence

A most possible sequence of status for success audio import

7. Advanced Functions

Besides the above simplified of the API calls, nexiwave provides advanced version of API calls. These APIs are meant to expose the full functionalities of nexiwave underlying system. It is recommended to try out the simplified version first.

7.1 Web Service: add (Full Parameter List)

You can supply more parameters for your upload.

Request:

*Parameter*	*Value*	*Size*	*Mandatory*	*Comments*
authData	component		Yes	See “Common Web Service Data” section
data	component
mediaURL	Strng	5000	Yes (unless mediaFileData present)	An URL (http/https, ftp, please)
mediaFileData	attachment		Yes (unless mediaURL present)	Base64 media data in the attachment. Please refer SOAP Attachment for how to attach binary file in SOAP call. Note: · This is NOT a recommended way for sending large media content to nexiwave. · Nexiwave recommends using the mediaURL parameter.
externalKey	string	5000		The external key from your system. Note: nexiwave simply records this data. No validation will be enforced to ensure the uniqueness of this key.
keywords	String	5000		A list of keywords or phrases to be used in language model adaptation. You may also pass in time coded keywords/phrases, using the “Nexiwave Human Text Convention”. Excellent accuracy improvement is reasonable. Integration Suggestions: · One method of obtaining these keywords and time codes is to use Video OCR service. · You may also extract all words/phrases from an associated PowerPoint presentation.
displayTitle	String	5000		An optional field for display in nexiwave Web UI.
notes	String	5000		Any notes
sourceKey	String	128		Your key to help Nexiwave separate content during: · SpeakerId process (SpeakerId accuracy may deteriorate if one account has too many potential speakers). · Searches keywords will not be applied to recordings with different sourceKey. For example, if your application has many users of its own, you may use this field to let Nexiwave builds separate Speaker database for each key you used. Ideally, this key should be consistent for your users.
mediaTypeSuffix	String	100		The value is generally the typical file suffix for your media file. For example, “mp3” for mp3 audio files, “wmv” for windows media file. Use this field to help our media decoder to identify the media content type.
humanTranscription	String	5000		The human transcribed text, if available. The human text will automatically enter nexiwave time stamping process. The time stamp results can be obtained using getTimeStampedHumanTranscription call. This is equivalent to a separate call of attachHumanTranscription()
humanTranscriptionFile	Bytes (Base64 encoded)			This is equivalent to a separate call of attachHumanTranscriptionFile() Note: If both humanTranscription or humanTranscriptionFile are present in the parameters, humanTranscription will be used.
decodingProfile	String	100		A case-sensitive valid decoding profile name. If no value, or invalid value, is passed in, a nexiwave configured default decoding profile will be used. You may use “listDecodingProfiles” API call to obtain a list of supported decoding profiles.
contextualDocURLs	String	5000		URLs for contextual documents. Separated by semi-colon (;). Document(s) will be downloaded and keywords/phrases extracted and used during indexing process.
contextualDocumentFileData	Byte[]			Contextual document that is associated with the media. Supported file types: · All Ms Office files: PowerPoint, Word, etc · PDF · HTML/XML · TXT Words and phrases will be utilized during indexing process.
contextualDocumentFileData2	Byte[]			-
contextualDocumentFileData3	Byte[]			-
contextualDocumentFileData4	Byte[]			-
contextualDocumentFileData5	Byte[]			-
httpNotifyBackURL	String	5000		Instead of your code actively polling for processing status, nexiwave can push the status to your system. A HTTP(S) GET url that will be called by nexiwave upon each status change event during the processing of your recording. Three extra parameters will be added to the url: “recordingId”, “externalKey” and “status” You may also configure the default value for this field in your nexiwave account “Settings” page, or through the Nexiwave profile API. Please read the below section “Notes on httpNotifyBackURL” for more information.
actualStartTimeInMS	long			The actual start time of the audio, in number of milliseconds after January 1^st, 1970. (as defined java.util.Date.getTime() method). If this field is not set, the current time at the upload will be used.
customIndexedProperties	Two dimension String array, Key/value pairs			These are the custom indexed properties, key/value pairs, for this particular recording. See below section “More information on customIndexProperties” for more information on this field.
processingPriority	int			Your processing priority. A recording with higher value of this attribute will be processed earlier than the one with lower value. Note: this processing priority is only effective within your account.
storeAudio	boolean			A Boolean field to indicate whether nexiwave should store the audio after processing. Default is true. Note: · If the value is false, all related audio files will be deleted upon finishing decoding. · Other artefacts, such as database record, generated text and timing info, will still remain in nexiwave system. These data may still be needed for other API calls, such as getTranscription. · A subsequent “remove” API call can be used to thoroughly delete everything related to this recording.
wordSpottingList	String	5000		List of words, separated by whitespace. Upon decoding finish, if any of the words were found in the audio, a HTTP notification will be sent back to your server, with a particular status of “WORD_SPOTTING_LIST_MATCHED”. For more information, see “Note on httpNotifyBackURL”.
otherAudioTrunks	component			If you record audio in separate file for each audio leg, you may send the audio leg separately for best accuracy.
mediaFile2TranscribeURL	String			The source URL for this audio leg/trunk
offsetInMS	long			The relative offset of the audio trunk to the start of the whole recording. Note: this field must be positive number. The behaviour is undefined for negative numbers.
hints	component
minNumberOfSpeakers	Int			Minimum number of speakers in the audio. Default is 1.
maxNumberOfSpeakers	Int			Maximum number of speakers. Zero or negative number means unlimited. Default is zero (unlimited)
speakerNames	String	1000		The speaker names, separated by semi-colon (;)

RESTFul GET Request:

https://api.nexiwave.com/services/SpeechIndexing/add?authData.email=XXX&authData.passwd=XXX&data.mediaURL=XXX&data.targetDecodingConfigName=XXX&....

Response:

Parameter	Value	Mandatory	Comments
Recording
recordingId	long	Yes

7.1.1 Sample Web Service Request and Response:

Sample Response

http://www.nexiwave.com/static/api/captured.soap.messages/addExt.response.xml

7.1.2 Notes on “httpNotifyBackURL”

Nexiwave can notify you about the status change of nearly each step of the processing, in an asynchronously manner.

Three parameters will be added to this GET url:

recordingId: this is the nexiwave internal recording ID assigned to your recording
externalKey: the externalKey that the particular recording was associated with.
status: this is the same status string you can obtain from the “queryProcessingStatus” request. Obviously, some statuses are transitional, which you may only receive in this notification back to you. for a list of possible status valid, please reference table “Possible Processing Status Values”

Example:

If the supplied notification URL was:

http://your.domain.com/app/callback

Then, the GET request that your server will receive will be:

http://your.domain.com/app/callback?recordingId=12345&status=WRITING_RECORDING

Your Nexiwave account can also be configured with the default notify back URL through the “Settings” page.

7.1.3 Notes on “customIndexedProperties”

You may associate one or more custom properties to a recording. This is a hash like properties.

Sample REST URL with customIndexedProperties:

https://api.nexiwave.com/services/SpeechIndexing/add?authData.email=XXX&authData.passwd=XXX&data.mediaURL=XXX&data.customIndexedProperties[myCustomerId]=1234&data.customIndexedProperties[receivedFromChannel]=XXX&…

Nexiwave may use these custom properties to better categorize your audio and process with proper models. Please contact support@nexiwave.com for setup.

7.2 Advanced Web Service: updateRecording

You may use this call to update some fields of a particular recording in the nexiwave system.

Request:

Parameter	Value	Size	Mandatory	Comments
authData	component		Yes	See “Common Web Service Data” section
recordingId	Long		Yes	The nexiwave recordingId
indexedAudioData	Component			This field is corresponding to the “addExt” field. Please refer to “addExt” for this field.
externalKey	String
mediaFileKeywords	String
displayTitle	String
notes	String
actualStartTimeInMS	Long
customIndexedProperties	Two dimension string array, key/value pairs

RESTFul GET Request:

https://api.nexiwave.com/services/SpeechIndexing/updateRecording?authData.email=XXX&authData.passwd=XXX&recordingId=XXX&indexedAudioData.externalKey=XXX&...

Response:

Parameter	Value	Mandatory	Comments
Result	boolean		Whether the update was success or not.

7.3 Other Advanced Web Service:

For performance reasons, nexiwave also offers batched requests for nearly all frequently used API calls.

Batch Request	Singular request	Comments
multiAddExt	addExt
multiQueryProcessingStatus	queryProcessingStatus
multiUpdateRecording	updateRecording

Comparing to the corresponding singular requests, batched requests accepts an array of input data and return an array of result data. Please refer the corresponding singular requests for details of the fields.

8. ProfileService

Nexiwave provides ProfileService at this location:

WSDL: http://api.nexiwave.com/services/ProfileService?wsdl

8.1.1 Web Service: changePasswd

You may change password of a nexiwave account.

Request:

Parameter	Value	Size	Mandatory	Comments
authData	component		Yes	See “Common Web Service Data” section
newPasswd	String	256	Yes	The new password

RESTFul GET Request:

https://api.nexiwave.com/services/ProfileService/changePasswd?authData.email=XXX&authData.passwd=XXX&newPasswd=XXX

Response:

Parameter	Value	Comments
result	boolean	Success or not

Please also check with Password Policy section.

8.1.2 Web Service: setDefaultHttpNotifyBackURL

You may use this API to set the default “HTTP Notify Back URL” for an account.

Parameter	Value	Size	Mandatory	Comments
authData	component		Yes	See “Common Web Service Data” section
defaultHttpNotifyBackURL	String	5000	Yes	The new default “Http Notify Back URL”

RESTFul GET Request:

https://api.nexiwave.com/services/ProfileService/setDefaultHttpNotifyBackURL?authData.email=XXX&authData.passwd=XXX&defaultHttpNotifyBackURL=XXX

Response:

Parameter	Value	Comments
result	boolean	Success or not

9. Appendix

9.1 Supported Input / Output Formats

9.1.1 Input Formats

Nexiwave supports virtually all kind of media format. For best accuracy, nexiwave recommends un-compressed PCM Wave audio.

Nexiwave is capable of extracting audio from these media file formats.

Supported video/audio codec:

(S)VCD (Super Video CD)
CDRwin's .bin image file
MPEG-1/2 (ES/PS/PES/VOB)
AVI file format
ASF/WMV/WMA format
QT/MOV/MP4 format
RealAudio/RealVideo format
Ogg/OGM files
Matroska
NUT
VIVO format
FLI format
NuppelVideo format
yuv4mpeg format
FILM (.cpk) format
RoQ format
PVA format

Supported audio codec:

MPEG layer 1, 2, and 3 (MP3) audio
AC3/A52, E-AC3, DTS (Dolby Digital) audio (software or SP/DIF)
AAC (MPEG-4 audio)
WMA (DivX Audio) v1, v2
WMA 9 (WMAv3), Voxware audio, ACELP.net etc (using x86 DLLs)
RealAudio: COOK, SIPRO, ATRAC3 (using Real libraries)
RealAudio: DNET and older codecs
QuickTime: Qclp, Q-Design QDMC/QDM2, MACE 3/6 (using QT libraries), ALAC
Ogg Vorbis audio
VIVO audio (g723, Vivo Siren) (using x86 DLL)
alaw/ulaw, (ms)gsm, pcm, *adpcm and other simple old audio formats
3gp, amr

9.1.2 Output Formats

Nexiwave can output machine processing results in these formats:

File type name	Content type	Description
timedtext.xml	text/xml	Timedtext for subtitle
sami.xml	text/xml	SAMI content for subtitle
qt.txt	text	QuickTime subtitle
caption.srt	text	Subrip SRT subtitle
info.txt	text	The.info format. See format definition here.
pure.text.txt	Text	Pure text. With all words in one line.

9.2 Audio File Format Recommendations

9.2.1 Record per speaker per channel

For best accuracy, audio files are processed on a per channel basis. It is recommended to record your audio source file on per speaker per channel basis.

9.2.2 Recording frequency

Nexiwave system is prepared to accept nearly any kind of sample rate. Your audio file is converted to a nearest sample rate that nexiwave can accept.

Human speaks at less than 11,000hz. Therefore, there is also no need to use too high sample rate.

9.3 Nexiwave User Password Policy

Password length must have six or more
Password must have at least one digit
Password must have at least one English letter

9.4 Error Handling

List of error codes that Nexiwave uses:

Error Code (All error codes start with “Error.”)	Description
unable.to.create.user.exists.need.email.verification	Unable to create a user. The email address has been register with an account already, which is still waiting for user activation.
unable.to.create.user.exists.email.verified	Unable to create a user. The email address has been register with an active account already.
registration.invalidEMail	Invalid email used for registration
registration.invalidPassword	Invalid password used for registration
readonly.account	While invoking a non-read-only action, this account is read-only.
profile.changepassword.newpassword.empty	New password is empty
profile.password.minimum.length	New password is too short. It must have the length of six or more.
profile.password.must.has.digit	The new password must have at least one digit.
profile.password.must.has.letter	The new password must have at least one letter.
nothing.to.upload	While add request, nothing to import.
speaker.ID.not.exist	While add request, speaker ID is specified, but not found in Nexiwave DB.
invalid.recordingId	While a recording access call, recordingId is not valid.
empty.query	Requested a search, the query is empty.
security.invalid.access.to.recording	Security error. The caller does not have access to the recording.
security.not.authorized.feature	Security error. The caller is not authorized to use the feature.
security.invalid.userId.and.password	Security error. Unable to authenticate with userId and password supplied.
security.email.not.verified	Security error. userId and password matches, but email has not been verified. The user should check verification email that was sent by Nexiwave.
security.account.locked.out	Security error. UserId and password matches, but the account has been locked out due to too many failed attempts. Please contact support@nexiwave.com to unlock the account.
security.invalid.request	Security error. The request is not valid.

Sample error message:

<soapenv:Fault>

<soapenv:Code>

<soapenv:Value>

axis2ns4:Error.security.invalid.userId.and.password

</soapenv:Value>

</soapenv:Code>

<soapenv:Reason>

<soapenv:Text xml:lang="en-US">

Unable to authenticate with supplied userid/passwd for user Auth Request for user test@nexiwave.com, group 0, passwd: (length=0). Result : FAILURE. Please contact support@nexiwave.com

</soapenv:Text>

</soapenv:Reason>

<soapenv:Detail/>

</soapenv:Fault>

9.5 .info file format

Nexiwave supports exporting speech processing results in .info format. A file of this format is a plain text file. Each line corresponds to a single word in the automatic transcription. Each line contains these fields, separated by one or more spaces:

<word> <confidence score:[0-1000]> <start> <end> 0.0 <side:[1,2,…]>

word: the word recognized by the Nexiwave speech engine
confidence: 0~1000, the confidence given by Nexiwave speech engine.
start: the start time of this word, in milliseconds
end: the end time of this word, in milliseconds

Example:

HELLO	900	1770	2210	0.0	1
WELCOME	1000	2750	3170	0.0	2