CN114338617A - Audio and video auditing method and illegal number identification method based on video call - Google Patents
Audio and video auditing method and illegal number identification method based on video call Download PDFInfo
- Publication number
- CN114338617A CN114338617A CN202111590933.7A CN202111590933A CN114338617A CN 114338617 A CN114338617 A CN 114338617A CN 202111590933 A CN202111590933 A CN 202111590933A CN 114338617 A CN114338617 A CN 114338617A
- Authority
- CN
- China
- Prior art keywords
- sip session
- audio
- video
- media stream
- video media
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 230000011664 signaling Effects 0.000 claims abstract description 35
- 238000007405 data analysis Methods 0.000 claims abstract description 31
- 238000004458 analytical method Methods 0.000 claims abstract description 18
- 239000000284 extract Substances 0.000 claims abstract description 6
- 238000007726 management method Methods 0.000 claims description 58
- 238000012552 review Methods 0.000 claims description 22
- 230000006399 behavior Effects 0.000 claims description 18
- 238000012795 verification Methods 0.000 claims description 12
- 238000012550 audit Methods 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 8
- 239000006185 dispersion Substances 0.000 claims description 8
- 238000005516 engineering process Methods 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 238000004891 communication Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 2
- 208000001613 Gambling Diseases 0.000 description 1
- 208000003028 Stuttering Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000391 smoking effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Images
Landscapes
- Telephonic Communication Services (AREA)
Abstract
The invention provides an audio and video auditing method and an illegal number identification method based on video calling, and the technical scheme is as follows: a call control platform acquires a call signaling message of an sip session from a telecommunication operator network, and extracts a sip session record from the call signaling message of the sip session; and when determining that audio and video auditing needs to be carried out on the sip session according to the sip session record, executing the following operations: sending the sip session record to a big data analysis platform for storage and call behavior analysis, and pushing the analyzed suspected illegal number to an operation management platform for manual rechecking; and acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform for auditing, and pushing the audio and video media stream information of the sip session record and the sip session to an operation management platform for manual rechecking when the audio and video media stream of the sip session is determined to be suspected to be illegal by auditing.
Description
Technical Field
The invention relates to the technical field of communication, in particular to an audio and video auditing method and an illegal number identification method based on video calling.
Background
With the development of communication technology, the demand of people for communication has been changed from a single voice demand to a comprehensive demand for video and audio. Video communication services that integrate voice, image, and video transmission are becoming popular communication services at present.
The video communication service enriches the lives of people and brings much trouble to people. Some lawbreakers apply video technology to carry out illegal advertising and spread bad contents related to yellow-relating politics and the like, and disturb normal social order. Telecommunication users want to have a safe conversation environment, and the occurrence of harassment, fraud and illegal advertisements is reduced and stopped. Government departments such as national security departments and public security, telecom operators and the like also hope to discover and dispose from sources through technical approaches, avoid the propagation of illegal and illegal contents, and provide a safe and credible communication environment for the public.
At present, the audio and video auditing method in video call mainly adopts manual auditing as main parts, has low timeliness and lagged processing, and because millions or even tens of millions of video contents can be generated in a network every day, if all the manual auditing is carried out, a large amount of manpower and time cost are consumed, and the large-scale full coverage cannot be realized by adopting a spot inspection mode.
Disclosure of Invention
In view of this, the present invention provides an audio and video auditing method and an illegal number identification method based on video call, so as to improve the timeliness and coverage of auditing and reduce auditing cost.
In order to achieve the purpose, the invention provides the following technical scheme:
an audio and video auditing method based on video calling is applied to a call control platform, and comprises the following steps:
acquiring a call signaling message of an sip session from a telecommunication operator network, and extracting a sip session record from the call signaling message of the sip session;
and when determining that audio and video auditing needs to be carried out on the sip session according to the sip session record, executing the following operations:
sending the sip session record to a big data analysis platform for storage, so that the big data analysis platform periodically performs call behavior analysis on the stored sip session record, and pushes the analyzed suspected illegal number to an operation management platform for manual review;
and acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform so that the content auditing platform audits the audio and video media stream of the sip session, and pushing the audio and video media stream information of the sip session record and the sip session to an operation management platform for manual rechecking when the auditing confirms that the audio and video media stream of the sip session is suspected to be illegal.
An audio and video auditing method based on video calling is applied to a content auditing platform, and comprises the following steps:
receiving an sip session record and an audio/video media stream of an sip session pushed by a call control platform, wherein the sip session record is obtained by extracting a call signaling message of the sip session from the call signaling message of the sip session after the call control platform acquires the call signaling message of the sip session from a telecommunication operator network; the audio and video media stream of the sip session is obtained from a telecommunication operator network after a call control platform determines that audio and video verification needs to be carried out on the sip session according to the sip session record; the call control platform acquires the audio and video media stream of the sip session from the telecom operator network and then pushes the sip session record and the audio and video media stream of the sip session to the content auditing platform;
performing voice recognition on the audio and video media stream of the sip session, and if the voice recognition result shows that the audio and video media stream of the sip session comprises forbidden words or forbidden voices, determining that the audio and video media stream of the sip session is suspected to be illegal;
carrying out picture identification on the audio and video media stream of the sip session, and if the picture identification result shows that the audio and video media stream of the sip session comprises forbidden pictures, determining that the audio and video media stream of the sip session is suspected to be illegal;
after determining that the audio/video media stream of the sip session is suspected to be illegal, pushing the sip session record and the audio/video media stream information of the sip session to an operation management platform for manual rechecking, and after the manual rechecking confirms that the audio/video media stream of the sip session is illegal, sending a calling number in the sip session record as an illegal number to a call control platform by the operation management platform, so that the call control platform updates a pre-configured calling blacklist according to the illegal number.
An illegal number identification method is applied to a big data analysis platform and comprises the following steps:
receiving and storing a sip session record pushed by a call control platform, wherein the sip session record is obtained by extracting a call signaling message of a sip session from the call signaling message of the sip session after the call control platform acquires the call signaling message of the sip session from a telecommunication operator network, and pushing the sip session record to a big data analysis platform after determining that audio and video verification needs to be carried out on the sip session according to the sip session record;
and regularly analyzing the calling behavior of the stored sip session record, pushing the analyzed illegal numbers to an operation management platform for manual rechecking, and sending the illegal numbers confirmed after manual rechecking to the calling control platform by the operation management platform, so that the calling control platform updates the pre-configured calling blacklist according to the illegal numbers confirmed by manual rechecking.
According to the technical scheme, after the call control platform acquires the call signaling message of the sip session from the telecommunication operator network and extracts the sip session record, the call control platform sends the sip session record to the big data analysis platform for call behavior analysis under the condition that audio and video verification needs to be carried out on the sip session according to the sip session record, and pushes the suspected illegal number to the operation control platform for manual recheck; and acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform for auditing, and pushing the audio and video media stream information of the sip session record and the sip session to an operation management platform for manual rechecking when the audio and video media stream of the sip session is determined to be suspected to be illegal by auditing. Therefore, by auditing the big data analysis platform and the content auditing platform, the sip sessions without suspected illegal information can be filtered out, so that the manual auditing can be performed only on the unfiltered sip sessions, the auditing timeliness and the coverage range are improved, and the auditing cost can be greatly reduced.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive labor.
Fig. 1 is a schematic architecture diagram of an audio and video auditing system based on a video call according to an embodiment of the present invention;
fig. 2 is a flowchart of an audio and video auditing method based on video call according to an embodiment of the present invention;
fig. 3 is a flowchart of an audio/video auditing method based on video call according to a second embodiment of the present invention;
fig. 4 is a flowchart of an audio and video auditing method based on video call according to a third embodiment of the present invention;
FIG. 5 is a flowchart of an illegal number identification method according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are, for example, capable of operation in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprising" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements explicitly listed, but may include other steps or elements not explicitly listed or inherent to such process, method, article, or apparatus.
The technical solution of the present invention will be described in detail with specific examples. Several of the following embodiments may be combined with each other and some details of the same or similar concepts or processes may not be repeated in some embodiments.
Referring to fig. 1, fig. 1 is a schematic structural diagram of an audio/video auditing system based on a video call according to an embodiment of the present invention, and as shown in fig. 1, the system includes a call management and control platform 10, a big data analysis platform 20, a content auditing platform 30, and an operation management platform 40; wherein the content of the first and second substances,
the call control platform 10 is configured to obtain a call signaling message and an audio/video media stream of a sip session from a telecommunications operator network, extract a sip session record from the call signaling message of the sip session, send the sip session record to the big data analysis platform 20 when determining that audio/video verification needs to be performed on the sip session according to the sip session record and a pre-configured calling blacklist, calling white list, and called blacklist, and push the sip session record and the audio/video media stream of the sip session to the content verification platform 30;
the big data analysis platform 20 is used for periodically analyzing the calling behavior of the sip session records received and stored from the call control platform 10 and sending the suspected illegal numbers analyzed to the operation management platform 40 for manual review;
the content auditing platform 30 is used for performing audio and video auditing on the audio and video media stream of the sip session pushed by the call control platform 10, and sending the audio and video media stream information of the sip session and the sip session record to the operation management platform 40 for manual review when the audio and video media stream of the sip session is determined to be illegal through auditing;
the operation management platform 40 is used for manually rechecking the suspected illegal numbers sent by the big data analysis platform 20, confirming the suspected illegal numbers as illegal numbers through manual rechecking and then sending the illegal numbers to the call control platform 10, so that the call control platform 10 updates the calling blacklist; the system is used for manually rechecking the audio and video media stream of the sip session according to the audio and video media stream information of the sip session pushed by the content auditing platform 30, and after the audio and video media stream of the sip session is confirmed to be illegal through manual rechecking, the calling number of the sip session is used as an illegal number and sent to the call control platform 10, so that the call control platform 10 updates the calling blacklist.
Referring to fig. 2, fig. 2 is a flowchart of an audio and video auditing method based on a video call according to an embodiment of the present invention, where the method is applied to a call management and control platform, and as shown in fig. 2, the method mainly includes the following steps:
and 204, acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform, so that the content auditing platform audits the audio and video media stream of the sip session, and pushing the audio and video media stream information of the sip session record and the sip session to an operation management platform for manual rechecking when the auditing confirms that the audio and video media stream of the sip session is suspected to be illegal.
As can be seen from the method shown in fig. 2, in this embodiment, after the call management and control platform obtains the call signaling message of the sip session from the telecom operator network and extracts the sip session record, under the condition that it is determined that audio and video auditing is required for the sip session according to the sip session record, the sip session record is sent to the big data analysis platform for storage and call behavior analysis, and the analyzed suspected illegal number is pushed to the operation management platform for manual review; and acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform for auditing, and pushing the audio and video media stream information of the sip session record and the sip session to an operation management platform for manual rechecking when the audio and video media stream of the sip session is determined to be suspected to be illegal by auditing. The big data analysis platform and the content auditing platform can filter the sip sessions without suspected illegal information through illegal number analysis and content auditing of the audio and video media stream, so that the unfiltered sip sessions can be manually audited, the sip sessions without suspected illegal information are subjected to spot inspection, the auditing timeliness and coverage range can be improved, and the auditing cost can be greatly reduced.
Referring to fig. 3, fig. 3 is a flowchart of an audio and video auditing method based on a video call according to a second embodiment of the present invention, where the method is applied to a call management and control platform, and as shown in fig. 3, the method mainly includes the following steps:
step 301, obtaining a call signaling message of an sip session from a telecommunication operator network, and extracting a sip session record from the call signaling message of the sip session;
in this embodiment, the telecom operator network may identify the call signaling message and the audio-video media stream of the sip session by using a traffic egress location deployed in the telecom operator network.
In this embodiment, the extracting the sip session record from the call signaling message of the sip session may specifically include the following contents: calling number, called number, conversation time interval (including starting time and ending time), conversation duration, a connection mark for indicating whether the call is connected, conversation type and other information, wherein the calling type comprises: video calls, audio calls, etc.
in this step 3021, determining whether interception of the sip session is required according to the sip session record and a pre-configured calling blacklist and a pre-configured called blacklist, which may specifically include the following steps:
if the calling number of the sip session is found in a pre-configured calling blacklist or the called number of the sip session is found in a pre-configured called blacklist, determining that the sip session needs to be intercepted, otherwise, determining that the sip session does not need to be intercepted.
In practical implementation, after determining that the sip session needs to be intercepted, the sip session may be further intercepted, so that the sip session is terminated, and after determining that the sip session does not need to be intercepted, the sip session may be released, so that the sip session is normally performed.
Step 3024, determining that audio and video auditing needs to be performed on the sip session, and turning to step 303 and step 304;
in this embodiment, if it is determined that audio/video auditing needs to be performed on a sip session, audio/video acquisition may be performed on the sip session to obtain audio/video media stream data (including an audio file and a video file) of the sip session.
And step 3025, determining that audio and video auditing of the sip session is not required.
In this embodiment, audio and video auditing is performed only on sip sessions in which the calling number does not belong to the calling blacklist or the calling white list, the called number does not belong to the called blacklist, and the call type is a video call.
The above steps 3021 to 3025 are specific refinements of step 202 shown in fig. 2.
in this embodiment, when the analyzed suspected illegal number is pushed to the operation management platform for manual rechecking, the analyzed suspected illegal reason can be simultaneously pushed to the operation management platform for reference of the manual rechecking.
And 304, acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform, so that the content auditing platform audits the audio and video media stream of the sip session, and pushing the audio and video media stream information of the sip session record and the sip session to an operation management platform for manual rechecking when the auditing confirms that the audio and video media stream of the sip session is suspected to be illegal.
In this embodiment, after determining that audio/video auditing needs to be performed on the sip session, the MGW/SBC may be controlled to initiate media renegotiation by modifying the SDP message body, and the calling and called media are bridged to the media server, thereby completing the media streaming function. The media server can sample and store the audio and video stream in the call. Each sip session will generate an audio file and a video file, and the audio file and the video file (i.e. audio/video media stream of the sip session) generated by the sip session can be named according to a uniform rule and automatically stored in the file server for the offline analysis of the content auditing platform. The whole process does not influence the normal operation of the sip session, and the client does not sense the session.
In practical application, after manual review determines that a suspected illegal number sent by the big data analysis platform is an illegal number, the illegal number can be sent to the call control platform, so that the call control platform incorporates the illegal number into the calling blacklist, and a subsequent call control platform can intercept a corresponding sip session when acquiring a signaling message of the sip session using the illegal number as the calling number again. Similarly, when the audio/video media stream of the suspected illegal sip session sent by the content auditing platform is confirmed to be the illegal audio/video media stream through manual review, the calling number of the sip session can be used as an illegal number to be sent to the call control platform, so that the call control platform can merge the illegal number into the calling blacklist, and the subsequent call control platform can intercept the corresponding sip session when acquiring the signaling message of the sip session using the illegal number as the calling number again.
Therefore, the present embodiment may further include the following step 305:
305, receiving an illegal number sent by the operation management platform, and updating a pre-configured calling blacklist according to the received illegal number; the illegal number is an illegal number which is received by the operation management platform from a big data analysis platform and confirmed by manual review, or the operation management platform receives the sip session record and the audio and video media stream information of the sip session from a content verification platform, confirms that the audio and video media stream of the sip session is illegal by manual review, and extracts a calling number which is used as the illegal number from the received sip session record.
In practical application, the call limit time, for example, 1 month of call limit or permanent call limit, may be set for the mobile phone numbers in the calling blacklist and the called blacklist. After some illegal number is added into the calling blacklist and the corresponding call limit time is set, all sip sessions taking the mobile phone number as the calling number are intercepted within the call limit time of the mobile phone number, and the mobile phone number can be deleted from the calling blacklist after the call limit time of the mobile phone number is exceeded.
In addition, because new illegal numbers can be continuously confirmed in the operation management platform in a manual rechecking mode and can be added into the calling blacklist, so that the calling blacklist is continuously updated, the call management and control platform filters the sip session according to the calling blacklist, the called blacklist and the calling white list, the content auditing platform audits the unfiltered sip session with audio and video to find suspected illegal numbers, the operation management platform determines whether the suspected illegal numbers are really illegal numbers through manual rechecking and sends the confirmed illegal numbers to the call management and control platform, and the call management and control platform performs closed loop updating on the calling blacklist, so that the continuous operation of the whole system is ensured.
As can be seen from the method shown in fig. 3, in this embodiment, after obtaining a call signaling message of a sip session from a telecommunications operator network and extracting a sip session record, the call management and control platform 10 determines whether the sip session needs to be intercepted or released according to the sip session record, a calling blacklist, a calling white list, and a called blacklist, and when it is determined that the sip session does not need to be intercepted and released, if it is determined that the call type of the sip session is a video call, it is determined that audio and video auditing needs to be performed on the sip session, and further, the sip session record is sent to a big data analysis platform for storage and call behavior analysis, and the analyzed suspected illegal numbers are pushed to the operation management platform for manual rechecking, and the illegal numbers confirmed by manual service are sent to the call management and control platform for updating the calling blacklist; and acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform for auditing, pushing the sip session record and the audio and video media stream information of the sip session to an operation management platform for manual rechecking when the audio and video media stream of the sip session is determined to be suspected to be illegal through auditing, and sending the calling number of the sip session to which the suspected illegal audio and video media stream confirmed by manual service belongs to a call control platform as an illegal number for updating a calling blacklist. Because the big data analysis platform and the content auditing platform can filter out the sip sessions without suspected illegal information through illegal number analysis and content auditing of the audio and video media stream, the artificially auditing can be carried out only on the unfiltered sip sessions, thereby not only improving the timeliness and coverage range of the auditing, but also greatly reducing the auditing cost.
Referring to fig. 4, fig. 4 is a flowchart of an audio and video auditing method based on a video call according to a third embodiment of the present invention, where the method is applied to a content auditing platform, and as shown in fig. 4, the method mainly includes the following steps:
in this embodiment, the performing voice recognition on the audio/video media stream of the sip session specifically includes:
s21, converting the audio and video media stream of the sip session from voice to text, identifying forbidden words of text information obtained by voice conversion, and if the forbidden words are identified from the text obtained by voice conversion, determining that the audio and video media stream of the sip session comprises the forbidden words;
here, the forbidden words may be set in advance according to requirements, for example, some related words related to yellow, political, violent, and abuse may be added as forbidden words to the forbidden word library. And determining whether the audio and video media stream of the sip conversation comprises forbidden words or not by identifying whether the text obtained by voice conversion contains the forbidden words in the forbidden word library or not.
S22, performing tone recognition on the audio and video media stream of the sip session through a voiceprint detection method, and if forbidden voices are recognized, determining that the audio and video media stream of the sip session comprises the forbidden voices.
Here, the prohibited voice may be preset according to a requirement, for example, a voice segment of a cheerful voice may be used as the prohibited voice, and when it is detected by a voiceprint detection method that the audio/video media stream of the sip session includes a voice segment with a higher similarity to the prohibited voice, it may be determined that the audio/video media stream of the sip session includes the prohibited voice.
In addition, in this embodiment, voice quality detection may be performed on the audio/video media stream of the sip session, and when it is detected that the audio/video media stream of the sip session has audio quality problems such as stutter, silence, and no sound, it may also be determined that the audio/video media stream of the sip session is suspected to be illegal, so that the sip session record and the audio/video media stream information of the sip session are sent to an operation management platform for manual review.
in this embodiment, the picture recognition is performed on the audio/video media stream of the sip session, and information or content which is related to yellow, riot, advertisement, bad scenes and the like and possibly violated rules can be recognized at the maximum probability.
In this embodiment, the image recognition of the audio/video media stream of the sip session specifically includes:
s31, recognizing each frame of video picture of the audio and video media stream of the sip session by utilizing a pre-trained illegal picture recognition model, and if the audio and video media stream of the sip session is recognized and determined to comprise the illegal video picture, determining that the audio and video media stream of the sip session comprises an illegal picture;
here, a large number of illegal pictures, such as pictures related to yellow and riot terrorism, can be collected in advance, and the illegal pictures are trained by combining a neural network algorithm to obtain an illegal picture recognition model, so that the illegal video pictures can be recognized by using the illegal picture recognition model.
S32, detecting characters, watermarks, two-dimensional codes and small program codes in each frame of video pictures of the audio and video media stream of the sip session, and determining that the audio and video media stream of the sip session comprises forbidden pictures when determining that the video pictures contain graphic advertisements according to the characters, watermarks, two-dimensional codes or small program codes identified from the video pictures;
in practical application, an illegal organization person usually performs fraud on a user by sending an audio/video media stream containing advertisement information, so that whether a video picture contains a text advertisement or not can be determined by detecting characters, watermarks, two-dimensional codes and small program codes in each frame of video picture of the audio/video media stream of the sip session, and when the video picture contains the text advertisement, it can be determined that the audio/video media stream of the sip session contains forbidden pictures.
S33, identifying a trademark in each frame of video picture of the audio and video media stream of the sip session by using a target detection technology, and if the trademark forbidden to be used in the video call is identified, determining that the audio and video media stream of the sip session comprises a forbidden picture;
in practical application, some logos (trademarks/logos) are not allowed to be spread for copyright protection, so that trademark identification can be performed on each frame of video picture of the audio/video media stream of the sip session by using an object detection technology, and if the identified trademarks are the trademarks which are not allowed to be spread, it is determined that forbidden pictures are included in the audio/video media stream of the sip session.
S34, carrying out illegal scene recognition on the audio and video media stream of the sip conversation by utilizing behavior analysis and time sequence comparison technology, and if the illegal scene is recognized, determining that the audio and video media stream of the sip conversation comprises forbidden pictures.
In practical application, an illegal scene recognition may be performed on the audio/video media stream of the sip session by using a behavior analysis and time sequence comparison technique, where the illegal scene may include: picture-in-picture, smoking, gambling, etc.
In addition, in this embodiment, picture quality detection may be performed on the audio/video media stream of the sip session, and when it is detected that there are picture quality problems of the audio/video media stream of the sip session, such as jittering ghosting, blurring, low illumination, overexposure, black screen, white screen, noise, mosaic, snow, static frame, frame skipping, etc., it may also be determined that the audio/video media stream of the sip session is suspected to be illegal, so that the sip session record and the audio/video media stream information of the sip session are sent to an operation management platform for manual review.
The above steps 402 and 403 may be executed in parallel or in series.
In practical implementation, the audio and video media stream (including audio file and video file) of the sip session may be stored in a database or other storage media, and the audio and video media stream information of the sip session may include file names and file storage locations of the audio file and the video file of the sip session, so that an auditor may find the audio file and the video file of the corresponding file storage location and perform manual review.
In actual implementation, an auditor can log in to the operation management platform to check the audit tasks assigned to the auditor. The auditing task comprises auditing records generated by the big data analysis platform. The audit record generated by the big data analysis platform may be empty, or may include one or more suspected illegal numbers and information such as an illegal reason of each suspected illegal number. The auditor can manually recheck each suspected illegal number.
As can be seen from the method shown in fig. 4, in this embodiment, the content auditing platform 30 performs audio and video auditing on the audio and video media stream of the sip session sent by the call control platform, and after determining that the audio and video media stream of the sip session is suspected to be illegal through auditing, sends the sip session record and the audio and video media stream information of the sip session to the operation management platform for manual review, and if it is determined that the audio and video media stream of the sip session is illegal through manual review, sends the calling number of the sip session as an illegal number to the call control platform, so that the call control platform updates a pre-configured calling blacklist according to the illegal number. Because the content auditing platform can filter the audio and video media stream which does not contain illegal content, the audio and video media stream containing illegal content only needs to be manually rechecked, the manual auditing cost can be greatly reduced, and the audio and video auditing of the audio and video media stream of the sip session can be performed through the content auditing platform, so that the timeliness and the coverage range of the auditing can be improved.
Referring to fig. 5, fig. 5 is a flowchart of an illegal number identification method according to an embodiment of the present invention, where the method is applied to a big data analysis platform, and as shown in fig. 5, the method mainly includes the following steps:
step 501, receiving and storing a sip session record pushed by a call control platform, wherein the sip session record is extracted from a call signaling message of a sip session after the call control platform acquires the call signaling message of the sip session from a telecommunication operator network, and is pushed to a big data analysis platform after determining that audio and video audit needs to be performed on the sip session according to the sip session record;
in this embodiment, the sip session record includes information such as a calling number, a called number, a session period (including a start time and an end time), a session duration, a connection flag indicating whether a call is connected, and a session type;
in this embodiment, the call behavior analysis is performed on all stored sip session records periodically, and includes:
s41, counting all sip session records in the latest preset counting duration, and determining the calling frequency, the dispersion of called numbers, the conversation frequency ratio of the conversation duration smaller than the preset conversation duration and the calling connection frequency ratio of each calling number in the latest preset duration;
in this embodiment, the number of sip session records of each calling number in the latest preset duration may be counted, one sip session record corresponds to one sip session, the number of sip session records of the calling number in unit time may be further determined, and the number of sip session records of the calling number in unit time is determined as the calling frequency of the calling number in the latest preset duration. The higher the frequency of calls, the greater the likelihood that the calling number is an illegal number.
In this embodiment, the number of sip session records of each calling number in the latest preset duration and the number of different called numbers included in the sip session records may be counted, and the number of the called numbers and the number of sip session records of the calling number in the latest preset duration are determined as the dispersion of the called numbers of the calling number. The higher the dispersion of the called number, the higher the possibility that the calling number is an illegal number.
In this embodiment, the number of sip session records of each calling number in the latest preset time duration may be counted, the number of sip session records in which the call time duration is less than the preset call time duration in the sip session records may be counted, and the ratio of the latter statistical result to the former statistical result is determined as the ratio of the call time duration to the call time duration less than the preset call time duration. The larger the proportion of the number of calls of which the call duration is less than the preset call duration is, the higher the possibility that the calling number is an illegal number is.
In this embodiment, the number of sip session records of each calling number in the latest preset time period may be counted, the number of sip session records in which calls are connected in the sip session records may be counted, and a ratio of a latter statistical result to a former statistical result is determined as a ratio of call connection times. The larger the call completion number ratio, the smaller the possibility that the calling number is an illegal number.
S42, determining the calling number with the calling frequency exceeding the preset calling frequency threshold value, the called number dispersion exceeding the preset dispersion threshold value, the calling time ratio of the calling time smaller than the preset calling time exceeding the preset first proportion and/or the calling connection time ratio smaller than the preset second proportion as a suspected illegal number.
In actual implementation, an auditor can log in to the operation management platform to check the audit tasks assigned to the auditor. The auditing task comprises auditing records generated by the content auditing platform. And each audit record corresponds to one call (one sip session) and comprises a calling number, a called number, a call time interval, abnormal reasons and the like, and supports audio and video playback, and an auditor can play back audio and video files on line and perform manual review.
In practical application, after the manual auditing task is assigned to the auditor, performance management can be performed on the auditor. The task completion condition of the auditor can be checked in real time through a WEB interface, and the task completion condition can show the audited and unchecked task details in a list form, so that the overall audit work progress can be conveniently controlled and the performance of the auditor can be conveniently checked. The task details comprise calling and called numbers, pre-review results, task assignment time, auditor information, review completion time, review results and the like.
As can be seen from the method shown in fig. 5, in this embodiment, the big data analysis platform periodically analyzes the sip session record pushed by the call management and control platform, so as to determine whether the calling number is illegal or not through analysis of the call behavior of each calling number within a period of time, and then sends the suspected illegal number to the operation management platform for manual review, so that the illegal number can be accurately identified, and the illegal number confirmed by the manual review is updated to the caller blacklist pre-configured by the call management and control platform, so that the call management and control platform can timely intercept the sip session dialed by the illegal number.
An embodiment of the present invention further provides an electronic device, as shown in fig. 6, where the electronic device includes: at least one processor 601, and a memory 602 connected to the at least one processor 601 through a bus; the memory 602 stores one or more computer programs executable by the at least one processor 601; the at least one processor, when executing the one or more computer programs, implements the steps in the method illustrated in any of the flowcharts of fig. 2-5.
Embodiments of the present invention also provide a computer-readable storage medium storing one or more computer programs which, when executed by a processor, implement the steps in the method shown in any of the flowcharts in fig. 2-5.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. An audio and video auditing method based on video calling is applied to a call control platform and is characterized by comprising the following steps:
acquiring a call signaling message of an sip session from a telecommunication operator network, and extracting a sip session record from the call signaling message of the sip session;
and when determining that audio and video auditing needs to be carried out on the sip session according to the sip session record, executing the following operations:
sending the sip session record to a big data analysis platform for storage, so that the big data analysis platform periodically performs call behavior analysis on the stored sip session record, and pushes the analyzed suspected illegal number to an operation management platform for manual review;
and acquiring the audio and video media stream of the sip session from a telecommunication operator network, sending the sip session record and the audio and video media stream of the sip session to a content auditing platform so that the content auditing platform audits the audio and video media stream of the sip session, and pushing the audio and video media stream information of the sip session record and the sip session to an operation management platform for manual rechecking when the auditing confirms that the audio and video media stream of the sip session is suspected to be illegal.
2. The method of claim 1,
the sip session record comprises a calling number, a called number and a calling type; the call types include audio calls and video calls;
determining whether audio and video auditing needs to be performed on the sip session according to the sip session record, wherein the determining comprises the following steps:
judging whether interception of the sip session is needed or not according to the sip session record and a pre-configured calling blacklist and a called blacklist;
if so, determining that audio and video auditing is not required for the sip session;
if not, further judging whether the calling number of the sip session is in a preset calling white list, if so, determining that audio and video verification is not needed to be performed on the sip session, and if not, determining that audio and video verification is needed to be performed on the sip session when the calling type of the sip session is a video call; and when the call type of the sip session is not a video call, determining that audio and video auditing on the sip session is not required.
3. The method of claim 2,
judging whether interception of the sip session is needed or not according to the sip session record, and a pre-configured calling blacklist and a pre-configured called blacklist, wherein the judging interception comprises the following steps:
if the calling number of the sip session is found in a pre-configured calling blacklist or the called number of the sip session is found in a pre-configured called blacklist, determining that the sip session needs to be intercepted, otherwise, determining that the sip session does not need to be intercepted.
4. The method of claim 2, further comprising:
receiving an illegal number sent by an operation management platform, and updating a pre-configured calling blacklist according to the received illegal number; the illegal number is an illegal number which is received by the operation management platform from the big data analysis platform and confirmed by manual rechecking, or the operation management platform receives the sip session record and the audio and video media stream information of the sip session from the content verification platform, confirms that the audio and video media stream of the sip session is illegal by manual rechecking, and extracts a calling number which is taken as the illegal number from the received sip session record.
5. An audio and video auditing method based on video calling is applied to a content auditing platform and is characterized in that the method comprises the following steps:
receiving an sip session record and an audio/video media stream of an sip session pushed by a call control platform, wherein the sip session record is obtained by extracting a call signaling message of the sip session from the call signaling message of the sip session after the call control platform acquires the call signaling message of the sip session from a telecommunication operator network; the audio and video media stream of the sip session is obtained from a telecommunication operator network after a call control platform determines that audio and video verification needs to be carried out on the sip session according to the sip session record; the call control platform acquires the audio and video media stream of the sip session from the telecom operator network and then pushes the sip session record and the audio and video media stream of the sip session to the content auditing platform;
performing voice recognition on the audio and video media stream of the sip session, and if the voice recognition result shows that the audio and video media stream of the sip session comprises forbidden words or forbidden voices, determining that the audio and video media stream of the sip session is suspected to be illegal;
carrying out picture identification on the audio and video media stream of the sip session, and if the picture identification result shows that the audio and video media stream of the sip session comprises forbidden pictures, determining that the audio and video media stream of the sip session is suspected to be illegal;
after determining that the audio/video media stream of the sip session is suspected to be illegal, pushing the sip session record and the audio/video media stream information of the sip session to an operation management platform for manual rechecking, and after the manual rechecking confirms that the audio/video media stream of the sip session is illegal, sending a calling number in the sip session record as an illegal number to a call control platform by the operation management platform, so that the call control platform updates a pre-configured calling blacklist according to the illegal number.
6. The method of claim 5,
and performing voice recognition on the audio and video media stream of the sip session, wherein the voice recognition comprises the following steps:
performing voice-to-text conversion on the audio and video media stream of the sip session, performing forbidden word recognition on text information obtained by voice conversion, and if forbidden words are recognized from the text obtained by voice conversion, determining that the audio and video media stream of the sip session comprises the forbidden words;
and performing tone recognition on the audio and video media stream of the sip session through a voiceprint detection method, and if forbidden voices are recognized, determining that the audio and video media stream of the sip session comprises the forbidden voices.
7. The method of claim 5,
performing picture recognition on the audio and video media stream of the sip session, including:
identifying each frame of video picture of the audio and video media stream of the sip session by using a pre-trained illegal picture identification model, and if the audio and video media stream of the sip session is identified and determined to comprise an illegal video picture, determining that the audio and video media stream of the sip session comprises an illegal picture;
detecting characters, watermarks, two-dimensional codes and small program codes in each frame of video pictures of the audio and video media streams of the sip session, and determining that the audio and video media streams of the sip session comprise forbidden pictures when determining that the video pictures comprise graphic advertisements according to the characters, watermarks, two-dimensional codes or small program codes identified from the video pictures;
identifying a trademark in each frame of video picture of the audio and video media stream of the sip session by using a target detection technology, and if the trademark forbidden to be used in the video call is identified, determining that the audio and video media stream of the sip session comprises a forbidden picture;
and carrying out illegal scene recognition on the audio and video media stream of the sip session by utilizing a behavior analysis and time sequence comparison technology, and if an illegal scene is recognized, determining that the audio and video media stream of the sip session comprises forbidden pictures.
8. An illegal number identification method is applied to a big data analysis platform and is characterized by comprising the following steps:
receiving and storing a sip session record pushed by a call control platform, wherein the sip session record is obtained by extracting a call signaling message of a sip session from the call signaling message of the sip session after the call control platform acquires the call signaling message of the sip session from a telecommunication operator network, and pushing the sip session record to a big data analysis platform after determining that audio and video verification needs to be carried out on the sip session according to the sip session record;
and regularly analyzing the calling behavior of the stored sip session record, pushing the analyzed illegal numbers to an operation management platform for manual rechecking, and sending the illegal numbers confirmed after manual rechecking to the calling control platform by the operation management platform, so that the calling control platform updates the pre-configured calling blacklist according to the illegal numbers confirmed by manual rechecking.
9. The method of claim 8,
the sip session record comprises a calling number, a called number, a call duration and a connection mark for indicating whether a call is connected or not;
and periodically carrying out call behavior analysis on all stored sip session records, wherein the call behavior analysis comprises the following steps:
counting all sip session records in the latest preset counting time length, and determining the calling frequency, the dispersion of called numbers, the conversation time ratio of the conversation time length smaller than the preset conversation time length and the calling connection time ratio of each calling number in the latest preset counting time length;
and determining the calling number with the calling frequency exceeding a preset calling frequency threshold value, the dispersion of the called number exceeding a preset dispersion threshold value, the calling time ratio of the calling time smaller than the preset calling time exceeding a preset first proportion and/or the calling connection time ratio smaller than a preset second proportion as a suspected illegal number.
10. An electronic device, comprising: the system comprises at least one processor and a memory connected with the at least one processor through a bus; the memory stores one or more computer programs executable by the at least one processor; characterized in that the at least one processor, when executing the one or more computer programs, implements the steps in the method of any one of claims 1-4, or implements the steps in the method of any one of claims 5-7, or implements the steps in the method of any one of claims 8-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111590933.7A CN114338617A (en) | 2021-12-23 | 2021-12-23 | Audio and video auditing method and illegal number identification method based on video call |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111590933.7A CN114338617A (en) | 2021-12-23 | 2021-12-23 | Audio and video auditing method and illegal number identification method based on video call |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114338617A true CN114338617A (en) | 2022-04-12 |
Family
ID=81054414
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111590933.7A Pending CN114338617A (en) | 2021-12-23 | 2021-12-23 | Audio and video auditing method and illegal number identification method based on video call |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114338617A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116452836A (en) * | 2023-05-10 | 2023-07-18 | 武汉精阅数字传媒科技有限公司 | New media material content acquisition system based on image data processing |
| CN117858079A (en) * | 2024-03-07 | 2024-04-09 | 四川众璟建设工程有限公司 | Safety control method of building intercom system |
| CN117858079B (en) * | 2024-03-07 | 2024-05-31 | 四川众璟建设工程有限公司 | Safety control method of building intercom system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104580649A (en) * | 2013-10-15 | 2015-04-29 | 中国移动通信集团河南有限公司 | Method and system for checking automatic speech service content |
| KR20160011120A (en) * | 2014-07-21 | 2016-01-29 | 주식회사 나의가는길 | Real-time voice phishing call attention smartphone application service provider system and method for guiding |
| CN107529068A (en) * | 2016-06-21 | 2017-12-29 | 北京新岸线网络技术有限公司 | Video content discrimination method and system |
| CN111541818A (en) * | 2020-04-10 | 2020-08-14 | 贵州索讯科技有限公司 | Fraud prevention method for screening, classifying and intercepting suspicious numbers based on big data |
| CN112511696A (en) * | 2020-11-11 | 2021-03-16 | 上海欣方智能系统有限公司 | System and method for identifying bad content of call center AI engine |
| CN113132368A (en) * | 2021-04-12 | 2021-07-16 | 海南晨风科技有限公司 | Chat data auditing method and device and computer equipment |
-
2021
- 2021-12-23 CN CN202111590933.7A patent/CN114338617A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104580649A (en) * | 2013-10-15 | 2015-04-29 | 中国移动通信集团河南有限公司 | Method and system for checking automatic speech service content |
| KR20160011120A (en) * | 2014-07-21 | 2016-01-29 | 주식회사 나의가는길 | Real-time voice phishing call attention smartphone application service provider system and method for guiding |
| CN107529068A (en) * | 2016-06-21 | 2017-12-29 | 北京新岸线网络技术有限公司 | Video content discrimination method and system |
| CN111541818A (en) * | 2020-04-10 | 2020-08-14 | 贵州索讯科技有限公司 | Fraud prevention method for screening, classifying and intercepting suspicious numbers based on big data |
| CN112511696A (en) * | 2020-11-11 | 2021-03-16 | 上海欣方智能系统有限公司 | System and method for identifying bad content of call center AI engine |
| CN113132368A (en) * | 2021-04-12 | 2021-07-16 | 海南晨风科技有限公司 | Chat data auditing method and device and computer equipment |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116452836A (en) * | 2023-05-10 | 2023-07-18 | 武汉精阅数字传媒科技有限公司 | New media material content acquisition system based on image data processing |
| CN116452836B (en) * | 2023-05-10 | 2023-11-28 | 杭州元媒科技有限公司 | New media material content acquisition system based on image data processing |
| CN117858079A (en) * | 2024-03-07 | 2024-04-09 | 四川众璟建设工程有限公司 | Safety control method of building intercom system |
| CN117858079B (en) * | 2024-03-07 | 2024-05-31 | 四川众璟建设工程有限公司 | Safety control method of building intercom system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109600752B (en) | Deep clustering fraud detection method and device | |
| US9531873B2 (en) | System, method and apparatus for classifying communications in a communications system | |
| US20190325345A1 (en) | Bot-based data collection for detecting phone solicitations | |
| US9571652B1 (en) | Enhanced diarization systems, media and methods of use | |
| US8391445B2 (en) | Caller identification using voice recognition | |
| CN108243049B (en) | Telecommunication fraud identification method and device | |
| CN106970911A (en) | A kind of strick precaution telecommunication fraud system and method based on big data and machine learning | |
| CN103095889A (en) | Junk call intercepting system based on talk mode identification and operating method thereof | |
| CN103731832A (en) | System and method for preventing phone and short message frauds | |
| CN106686191A (en) | Processing method for adaptively identifying harassing call and processing system thereof | |
| CN107770777B (en) | Method for identifying recorded fraud calls | |
| CN104639770A (en) | Telephone reporting method, device and system based on mobile terminal | |
| CA3057037A1 (en) | Systems and methods for anomaly or fraud detection | |
| CN112511696A (en) | System and method for identifying bad content of call center AI engine | |
| CN114338617A (en) | Audio and video auditing method and illegal number identification method based on video call | |
| CN103078984A (en) | Dialing method based on OCR (Optical Character Recognition) | |
| US10701205B2 (en) | Communications network | |
| CN101072174A (en) | Tencent voice identifying method based on pay load deep detection and session correlating technology | |
| CN111918226A (en) | Real-time signaling-based method and device for analyzing international high-settlement embezzlement behavior | |
| KR101033870B1 (en) | Method and device for processing spam call | |
| Manunza et al. | Kerberos: A real-time fraud detection system for IMS-enabled VoIP networks | |
| US20220103680A1 (en) | System and method for classifying and handling voice over ip traffic | |
| CN114257688A (en) | Telephone fraud identification method and related device | |
| GB2560961A (en) | Communications Network | |
| KR20080019770A (en) | Method for transmitting and replying short message |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |