CN109831696A - Handle method, apparatus, electronic equipment and the storage medium of violation video content - Google Patents
Handle method, apparatus, electronic equipment and the storage medium of violation video content Download PDFInfo
- Publication number
- CN109831696A CN109831696A CN201811626292.4A CN201811626292A CN109831696A CN 109831696 A CN109831696 A CN 109831696A CN 201811626292 A CN201811626292 A CN 201811626292A CN 109831696 A CN109831696 A CN 109831696A
- Authority
- CN
- China
- Prior art keywords
- time
- violation
- screenshot
- video content
- identification method
- 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.)
- Granted
Links
Landscapes
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
This application discloses a kind of method, apparatus, electronic equipment and storage mediums for handling violation video content, this method include obtain at the first time, at the first time characterization with the first identification method identify from video content include at first along video playing progress violation information time；Obtained for the second time, the second time representation with the second identification method identify from video content include at first along video playing progress violation information time；The safe period of time at the first time between the second time is obtained, safe period of time, first time and the second time are based on, obtains the violation timeliness of the second identification method, default processing is carried out to video content according to the violation identification timeliness got.By the safe period of time between removal live streaming violation screenshot and the time of the live streaming violation screenshot of machine recognition not including violation information, obtain the timeliness of machine recognition violation, video content break according to violation identification timeliness and broadcasts or break processing, can reduce the exposure time of violation information.
Description
Technical field
This application involves network direct broadcasting technical fields, more particularly, to a kind of method for handling violation video content, dress
It sets, electronic equipment and storage medium.
Background technique
With the fast development of internet and internet video live broadcasting technology, facilitates more and more users and passed by video
Information is broadcast, user experience is enhanced, but some problems also occurs, for example, some criminals disseminate bad disobey by video
Information is advised, huge negative effect can be brought to society.For miscellaneous live content, if cannot fast and accurately area
It separates normal live streaming and is broadcast live in violation of rules and regulations, it will generate situations such as largely failing to report or report by mistake, reduce user experience.
Summary of the invention
In view of the above problems, it the method, apparatus of violation video content, electronic equipment is handled and deposits present applicant proposes a kind of
Storage media, to improve the above problem.
In a first aspect, the embodiment of the present application provides a kind of method for handling violation video content, this method comprises: obtaining
At the first time, it characterizes at the first time with what the first identification method identified from video content and includes at first along video playing progress
The time of violation information, the first identification method are identification method trusty；Obtained for the second time, the second time representation is with second
What identification method identified from video content includes the time of violation information, the second identification method along video playing progress at first
For identification method to be detected；The safe period of time at the first time between the second time is obtained, safe period of time characterization is at the first time
It does not include the period of violation information in video content between the second time；When based on safe period of time, first time and second
Between, the violation for obtaining the second identification method identifies timeliness；Video content is preset according to the violation identification timeliness got
Processing, to reduce the exposure time of video content.
Second aspect, the embodiment of the present application provide a kind of device for handling violation video content, which includes: first
Module is obtained, for obtaining first time, the edge view that first time characterization is identified from video content with the first identification method
Frequency playback progress includes the time of violation information at first, and the first identification method is identification method trusty；Second obtains module,
For obtaining for the second time, the second time representation with the second identification method identify from video content along video playing progress
It at first include the time of violation information, the second identification method is identification method to be detected；Third obtains module, for obtaining the
Safe period of time between one time and the second time, safe period of time characterization is at the first time in the video content between the second time
It does not include the period of violation information；4th obtains module, for being based on the safe period of time, the first time and described second
Time, the violation for obtaining second identification method identify timeliness；Processing module, for identifying timeliness according to the violation got
Default processing is carried out to video content, to reduce the exposure time of video content.
The third aspect, the embodiment of the present application provide a kind of electronic equipment, including memory and one or more processing
Device；One or more programs are stored in memory and are configured as being performed by one or more processors, one or more
Program is configured to carry out method described in above-mentioned first aspect.
Fourth aspect, the embodiment of the present application provide a kind of computer readable storage medium, computer readable storage medium
In be stored with program code, wherein method described in above-mentioned first aspect is executed when program code is run as processor.
A kind of method, apparatus, electronic equipment and storage medium handling violation video content provided by the embodiments of the present application,
It is related to network direct broadcasting technical field.This method is by obtaining at the first time, and characterization is with the first identification method from video at the first time
What is identified in content includes the time of violation information along video playing progress at first, and the first identification method is identification trusty
Mode；Then obtained for the second time, the second time representation is broadcast with what the second identification method identified from video content along video
Degree of putting into includes the time of violation information at first, and the second identification method is identification method to be detected；Obtain again at the first time with
Safe period of time between second time, it does not include disobeying that safe period of time, which characterizes in the video content between the second time at the first time,
Advise the period of information；It is then based on safe period of time, first time and the second time, when obtaining the violation identification of the second identification method
Effect；Default processing is carried out to video content according to the violation identification timeliness got, to reduce the exposure time of video content.
This method is disobeyed by obtaining the live streaming violation screenshot time and obtaining the live streaming violation screenshot time of machine recognition by removal live streaming
Do not include the safe period of time of violation information between rule screenshot time and the live streaming violation screenshot time of machine recognition, obtains machine knowledge
The timeliness of other violation carries out default processing to video content according to violation identification timeliness, can reduce in video content and believe in violation of rules and regulations
The exposure time of breath.
Detailed description of the invention
In order to more clearly explain the technical solutions in the embodiments of the present application, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, the drawings in the following description are only some examples of the present application, for
For those skilled in the art, without creative efforts, it can also be obtained according to these attached drawings other attached
Figure.
Fig. 1 shows a kind of application environment schematic diagram provided by the embodiments of the present application.
Fig. 2 shows the audit logics suitable for the method for processing violation video content provided by the embodiments of the present application to show
It is intended to.
Fig. 3 shows a kind of flow chart of the method for processing violation video content of one embodiment of the application offer.
Fig. 4 shows a kind of flow chart of the method for processing violation video content that another embodiment of the application provides.
Fig. 5 shows the flow chart of the method for the step S210 in Fig. 4.
Fig. 6 shows the flow chart of the method for the step S220 in Fig. 4.
Fig. 7 shows the detection logical schematic suitable for the embodiment of the present application violation timeliness.
Fig. 8 shows a kind of structural block diagram of device for handling violation video content provided by the embodiments of the present application.
Fig. 9 shows the structural block diagram of a kind of electronic equipment provided by the embodiments of the present application.
Figure 10 shows disobeying for saving or carrying realization according to the processing of the embodiment of the present application for the embodiment of the present application
Advise the storage unit of the program code of the method for video content.
Specific embodiment
In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application
Attached drawing, the technical scheme in the embodiment of the application is clearly and completely described.
With the rapid development of information technology, network direct broadcasting is with its novel form, the abundant in content extensive pass by people
Note, can improve the spread speed of information by network direct broadcasting with real-time delivery information, but also bring many negative shadows simultaneously
It rings.For network direct broadcasting since its threshold is low, more and more users can become the main broadcaster of network direct broadcasting platform, therefore network direct broadcasting
Information category emerge one after another, some criminals are disseminated for example using network direct broadcasting: pornographic, violence, sham publicity etc. are bad
Information has upset the normal life of people.
As a kind of mode, platform, which is broadcast live, can be carried out some violation informations being spread out in video content by machine
Audit, however the violation information of a live video is not necessarily continuously, disseminate in the live video of violation information there is also
Not content in violation of rules and regulations, and machine recognition come out the violation time be usually all later than live video content it is practical beginning violation when
Between, i.e. the recognition capability of machine is insufficient, cause machine recognition video content whether violation when there are time delay, recognition efficiencies
It is low, it is therefore desirable to the method that one kind can accurately measure the identification timeliness of machine recognition violation.
In view of the above-mentioned problems, inventor has found by long-term research, since a live streaming intercutting violation information to separated
Rule information is come out by machine recognition, and there is a time difference in centre, and the recognition capability of machine is stronger, this time difference is shorter.
In order to reduce exposure period of the violation information in live video, to reduce violation information bring adverse effect, further
The recognition capability of machine is improved, inventors have found that first barrier of the machine recognition as live streaming audit, it can be by filtering out
Safe sample (the safe sample in the embodiment of the present application between violation time started to the violation time started of machine recognition is broadcast live
Refer to the video content not comprising violation information), and the sample of " doubtful violation " that machine recognition obtains is sent to doing some training very often
People examine (i.e. manual examination and verification) and carry out secondary audit, calculate the identification timeliness of machine recognition violation, it is possible to reduce violation information
Exposure time in video content further increases machine recognition ability.
Therefore, it is proposed to a kind of method, apparatus for handling violation video content provided by the embodiments of the present application, electronic equipment
And storage medium, by not including in violation of rules and regulations between removal live streaming violation screenshot time and the live streaming violation screenshot time of machine recognition
The safe period of time of information obtains the timeliness of machine recognition violation, demonstrates the ability of the timeliness of machine recognition violation；Furthermore root
Video content break according to violation identification timeliness and broadcasts or breaks processing, and then can reduce the exposure time of violation information.
For the ease of application scheme is described in detail, ring first is applied to one of the embodiment of the present application in conjunction with attached drawing below
Border is introduced.
Referring to Fig. 1, showing for a kind of application environment of method for handling violation video content provided by the embodiments of the present application
It is intended to, as shown in Figure 1, the application environment can be understood as a kind of network system 10 of the offer of the embodiment of the present application, the network
System 10 includes: server 11 and electronic equipment 12.Wherein, server 11 can be a server (network insertion service
Device), it is also possible to the server cluster (Cloud Server) consisted of several servers, or can be a cloud computing center
(database server).Electronic equipment 12 can be any equipment for having communication and store function, including but not limited to PC
(Personal Computer, personal computer), PDA (tablet computer), smart television, smart phone, intelligent wearable device
Or other intelligent communication devices with network connecting function.
Optionally, it is mounted with that the client can configured with the client of network direct broadcasting platform application in electronic equipment 12
To be application client (the video playing application program in such as cell phone application), it is also possible to webpage client (such as network
Platform is broadcast live).
As a kind of mode, server 11 can be established by network and electronic equipment 12 and be communicated to connect, optionally, the net
Network can be wireless network, be also possible to cable network.The user account that registration can be used in user logs in electronic equipment 12
Client or internet, the client can have information input interface, user inputs text information in the interface, and
And text information is shown in the chat interface of client.
It should be noted that method in the embodiment of the present application is applied to server, as a kind of mode, server can be with
It is a server 11 as shown in Figure 1, is also possible to the server being made of several servers 11 (only showing one in figure)
Cluster can install the live streaming auditing system of network direct broadcasting platform, be audited by the live streaming as a kind of mode in server 11
System can the video content to network direct broadcasting platform audited in real time.
In one embodiment, live streaming auditing system can be provided in advance comprising several violation pictures, text and/
Or the database of the classifications violation information such as video, these violation informations can include according to video content by live streaming auditing system
The violation severity of violation information is divided into different discipline ratings, then when recognizing in the video comprising any violation information
Rong Shi, live streaming auditing system can construct corresponding violation work order according to the discipline rating of video content, and violation work order is mentioned
It hands over the background management system to network direct broadcasting platform to be audited again, the accuracy of identification violation content can be increased.
The audit logic suitable for the method for processing violation video content provided by the embodiments of the present application is carried out below
Explanation.
Referring to Fig. 2, being patrolled for the audit suitable for the method for processing violation video content provided by the embodiments of the present application
Collect schematic diagram.As a kind of mode, machine can the content to live video screenshot was carried out with 3 seconds frequencies, and to each
The content of screenshot is identified judge whether it includes the violation information stored in the database for be broadcast live auditing system.In one kind
In implementation, when the content for identifying certain screenshot does not include above-mentioned violation information, live streaming auditing system can be by this
Screenshot is determined as " safe sample ", and should " safe sample " abandoned.In another implementation, when identifying certain
When the content of screenshot includes above-mentioned violation information, this screenshot can be determined as " doubtful violation " sample by live streaming auditing system,
And should " doubtful violation " sample send to people to examine and carry out secondary audit at (manual examination and verification).
As a kind of mode, it is sent to people and tries and punishes " doubtful violation " sample for carrying out secondary audit, if audit passes through, then
It can directly be abandoned corresponding screenshot as safe sample；If the audit fails, then can be true by corresponding screenshot
It is set to the screenshot comprising violation information, and it is punished.
For example, in a specific application scenarios, if by " doubtful violation " sample of machine recognition through artificial secondary careful
It is found after core not comprising violation information, then corresponding screenshot can be abandoned, i.e., is considered as the content of corresponding screenshot
Normal video content；If being found to include violation information after artificial secondary audit by " doubtful violation " sample of machine recognition,
So corresponding screenshot can be determined as in violation of rules and regulations by live streaming auditing system, i.e., the content of corresponding screenshot is considered as in improper video
Hold, in this case, live streaming auditing system can punish these video contents for including violation information, such as disconnected
It broadcasts, break；Or warning or fine etc..
It should be noted that above-mentioned " machine can content to live video with 3 seconds frequencies carry out screenshot " enumerated
Only to facilitate explanation, is not intended as limiting, for example, being also possible to machine in actual implementation to the content of live video with 10
The frequency of second carries out screenshot etc., does not repeat excessively herein.
It is understood that the audit logic signal in the method for processing violation video content provided by the embodiments of the present application
Figure is only for convenience of understanding, and the embodiment of the present application can be applicable in a variety of audit logics, therefore the side of above-mentioned processing violation video content
Audit logical schematic in method, which is not constituted this programme, to be limited.
Present embodiments are specifically described below in conjunction with attached drawing.
Referring to Fig. 3, a kind of flow chart of method for handling violation video content provided by the embodiments of the present application is shown,
The present embodiment provides a kind of methods for handling violation video content, can apply above-mentioned electronic equipment, this method comprises:
Step S110: obtaining at the first time, and the first time characterization is identified from video content with the first identification method
Out include at first along video playing progress violation information time, first identification method be identification method trusty.
Wherein, the first identification method in the present embodiment can be manual identified, be also possible to manual identified auxiliary machinery
Identification.It should be noted that the first identification method can be the identification method trusty by manual verification, it is also possible to lead to
Cross the identification method trusty of manual verification's auxiliary machinery verifying.As a kind of mode, auditing system is broadcast live can be with first
Whether identification method includes that violation information identifies, and then can will identify with first to the video content of network direct broadcasting platform
What mode identified from video content includes the time of violation information as at the first time along video playing progress at first.
As an implementation, if the first identification method is manual identified, then can be by manually in video content
Time addition comprising violation information marks in violation of rules and regulations, and using the time for occurring violation information at first along video playing progress as the
One time.By manual identified violation information, so mark time of violation at first as at the first time, can allow result more
Accurately.
As another embodiment, if the first identification method is the identification of manual identified auxiliary machinery, then can be by machine
Device is first classified to the video content of entire video by preconfigured violation information disaggregated model, obtains including in violation of rules and regulations
The video content of information and do not include violation information video content.Optionally, machine can after first subseries include in violation of rules and regulations
The video content of information is sent to people and tries and punishes carry out secondary-confirmation, by manually in the video including violation information of machine judgement
Appearance is judged again, at first includes in violation of rules and regulations when then people can be tried disconnected if it includes violation information that people's trial is disconnected optionally
The time of information is as at the first time；If it does not include violation information that people's trial is disconnected, then can using corresponding video content as
Safe sample is abandoned.Artificial secondary judgement is transferred to again by machine judgement, will examine the secondary packet at first for judging to obtain through people
The time of violation information is included as at the first time, it is ensured that when acquisition first time more quick and precisely, reduce error.
It should be noted that the manual identified and the identification of manual identified auxiliary machinery in the first identification method are credible
The identification method appointed, and two kinds of identification methods can obtain reliable first time.
Step S120: obtaining for the second time, and second time representation is identified from video content with the second identification method
Out include at first along video playing progress violation information time, second identification method be identification method to be detected.
Wherein, the second identification method in the present embodiment is machine recognition, as a kind of mode, the second identification method be to
The identification method of detection.Live streaming auditing system audit network direct broadcasting platform video content whether include violation information process
In, since the model of machine recognition is people's setting, the type of machine recognition may less comprehensively, when the view of network direct broadcasting platform
When occurring new (the machine recognition type being set is not covered with) violation information type in frequency content, machine recognition
Violation information is it is possible that situations such as failing to report, report by mistake or identify delay.
For example, being made whether by machine to video content in the review process comprising violation information, when in video content
When the violation information of appearance is not in the range of machine recognition model covers, machine will not recognize these new violation letters
Breath；Whether machine is during identifying video content includes violation information, due to some limitations etc. of machine learning algorithm
Reason can always have time delay, thus it is necessary to go acquisition machine recognition to go out the time that video content includes violation information.
As a kind of mode, auditing system is broadcast live can be with video content of second identification method to network direct broadcasting platform
It is no to be identified comprising violation information, so can by with the second identification method identify from video content along video playing
Progress includes the time of violation information as the second time at first.
Step S130: the safe period of time between the first time and second time, the safe period of time table are obtained
Levy do not include in the video content between the first time and second time violation information period.
As a kind of mode, edge in the video content identified due to machine recognition there are error, the first identification method
Video playing progress include at first in the first time of violation information and video content that machine recognition obtains along video playing into
Inevitably there is the safe period of time for not including violation information between the second time including violation information at first in degree, can also be with
It is interpreted as therebetween in the presence of the video content for not including violation information, then when live streaming auditing system can be by obtaining first
Between between the second time do not include violation information video content, and then between available first time and the second time
Safe period of time.
It should be noted that the second time in the embodiment of the present application is greater than at the first time.It can be appreciated that machine is known
Not Chu video content include violation information time be later than and actually violation information occur in video content at the beginning of.
Step S140: being based on the safe period of time, the first time and second time, obtains second identification
The violation of mode identifies timeliness.
As a kind of mode, the present embodiment can be used the second time, the difference with the sum of first time and safe period of time
It is worth and identifies timeliness as the violation of the second identification method.Specifically, including violation information since machine recognition goes out video content
Time is later than actually there is violation information in video content at the beginning of, and at the first time and the second time between exist do not wrap
The video content corresponding period of violation information is included, therefore, machine really identifies the timeliness of violation in order to obtain, can be by
Between one time and the second time do not include that the safe period of time of violation information is rejected, then, understandable to be, Ke Yitong
It crosses and knows the sum of the second time and safe period of time and first time as the violation that difference obtains the second identification method (i.e. machine recognition)
Other timeliness.
Optionally, " T2 " can be set by the value of the second time, sets " T1 " for the value of first time, when will be safe
The value of section is set as " T3 ", then the violation identification timeliness of above-mentioned second identification method can be expressed as " T2- (T1+T3) ", it can
Choosing, it is understood that be " T2-T1-T3 ".
For example, in a specific application scenarios, it is assumed that by first including believing in the video content of manual identified in violation of rules and regulations
The screenshot time of breath is video to the 5th second started is played, by first including violation information in the video content of machine recognition
The screenshot time be video to play start the 10th second, optionally, between the 5th second and the 10th second in the presence of 2 seconds do not include violation
The video content of information, then can be by the sum of 10 and 5 and 2 difference " 3 " disobeying as the second identification method (machine recognition)
Rule identification timeliness.
Step S150: carrying out default processing to the video content according to the violation identification timeliness got, to
Reduce the exposure time of the video content.
Wherein, as a kind of mode, if having got the violation identification timeliness of above-mentioned second identification method (machine recognition),
It so can be determined that there are time delays when the second identification method identifies the violation information of video content.Optionally, in order to avoid
Violation information brings adverse effect to people's lives, can be carried out according to the specific value of violation identification timeliness to video content
Default processing.
As an implementation, the threshold value for identifying timeliness in violation of rules and regulations can be preset, optionally, when violation identifies timeliness
When lower than the threshold value, the corresponding video content comprising violation information can slightly be punished, such as alert or penalize
Money etc.；When violation identification timeliness is not less than told threshold value, can break to the corresponding video content comprising violation information
Processing is broadcast or broken, so as to reduce the exposure time of violation information, promotes user experience.
A kind of method handling violation video content provided in this embodiment, by obtaining at the first time；Then is obtained
Two times；The safe period of time at the first time between the second time is obtained again；It is then based on safe period of time, first time and second
Time, the violation for obtaining the second identification method identify timeliness；Video content is carried out according to the violation identification timeliness got pre-
If processing, to reduce the exposure time of video content.This method is known by obtaining the live streaming violation screenshot time and obtaining machine
Other live streaming violation screenshot time, by between removal live streaming violation screenshot time and the live streaming violation screenshot time of machine recognition
Do not include the safe period of time of violation information, obtain the timeliness of machine recognition violation, demonstrates the energy of the timeliness of machine recognition violation
Power break to video content and broadcasts or break processing, can reduce the exposure time of violation information according to violation identification timeliness.
Referring to Fig. 4, showing a kind of stream of the method for processing violation video content that another embodiment of the application provides
Cheng Tu, the present embodiment provides a kind of methods for handling violation video content, can be applied to above-mentioned electronic equipment, this method comprises:
Step S210: obtaining first object unit screenshot, using the screenshot time of the first object unit screenshot as the
One time, the first object unit screenshot are to be identified from multiple unit screenshots that video content includes with the first identification method
Out include at first along video playing progress violation information unit screenshot.
Referring to Fig. 5, step S210 may include:
Step S211: the list that the violation label of the first identification method is carried in multiple unit screenshots of video content is obtained
First screenshot.
Optionally, for all video contents of network direct broadcasting platform, from the live streaming start and ending of video content, live streaming
Auditing system can carry out live video with preset frequency by machine cutting figure, obtain multiple video content screenshots.For example,
For one section of live video, auditing system is broadcast live can carry out cutting figure with content of 3 seconds frequency intervals to live video,
It can carry out cutting figure with content of 10 seconds frequency intervals to live video, obtain 3 seconds or 10 seconds comprising multiple video contents
Frequency screenshot.It therefore, can be using each screenshot in obtained multiple video content screenshots as one as a kind of mode
Unit screenshot.
As a kind of mode, network direct broadcasting platform can construct labeling system, which can be used for straight to network
All video content screenshot additions comprising violation information identify in violation of rules and regulations in broadcasting.It optionally, may include data in labeling system
Module is obtained, can be used for from internet grabbing and storing in real time includes but is not limited to relate to Huang, suspected illegal, bad advertisement etc.
The violation information of related category, wherein the format of violation information can be picture format, video/audio, text etc., herein not
It limits.It should be noted that the violation information in data acquisition module is also possible to by being manually entered, or manually to having deposited
The violation information that the violation information of storage obtains after modifying or adjusting.
As an implementation, above-mentioned labeling system can include different brackets according to the violation information building got
And include the violation mark of multiple classifications, for example, labeling system can be serious according to the violation of violation information as a kind of mode
Degree, setting " critical violation ", " generally in violation of rules and regulations ", " the violation mark of the three grades such as slight violation, for each violation etc.
Multiple violation classification logotypes can be respectively set in grade.
Alternatively, different violation marks can be arranged in labeling system according to the violation classification of violation information,
And the violation mark of each classification may include that one or more subclasses identify in violation of rules and regulations, such as " erotica content " is arranged, " violates method
Rule " etc. classification logotypes, for the classification logotypes such as " erotica content ", " contrary to law " can be respectively set it includes one or more
A sub- classification logotype optionally can also be arranged next stage classification logotype under subclass mark, can carry out according to the actual situation
Selection.There can be reference frame abundant in mark in order to auditor, can also be improved the justice to verifying video content
Property (avoid certain video contents including violation information due to identified category is very little labeled as in violation of rules and regulations or will actually including
The video content spill tag of violation information).
For example, in a specific application scenarios, if labeling system is distinguished according to the violation severity of violation information
Grade 1 violation mark provided with " critical violation ", " general violation ", " slight violation ", then " seriously being disobeyed for level-one mark
Rule ", it is " dressing is indecency " etc. that its second level sub-mark, which can be set,；For sub-mark " dressing is indecency ", its next stage can be set
Sub-mark (three-level mark) is " exposure should not naked position " etc..Optionally, the classification of mark and quantity can be according to realities
Border situation and be adjusted, details are not described herein.
In another specific application scenarios, if labeling system is according to the violation classification of violation information provided with " vulgar
Relate to Huang ", " porno advertisement ", the classifications such as " seriously relating to political affairs " violation mark can then be the mark of " vulgar to relate to Huang " for classification
It is " vulgar movement " etc. so that the sub-mark that it includes is arranged, is the mark of " porno advertisement " for classification, it, which can be set, includes
Sub-mark be " netpage color pop-up ", " the small card of contact method " etc., can be with for the marks that classification is " seriously relating to political affairs "
It is " personage " etc. that the sub-mark that it includes, which is arranged,.
Step S212: along video playing progress in the unit screenshot that the violation for carrying the first identification method is marked
The unit screenshot occurred earliest is as first object unit screenshot, using the screenshot time of the first object unit screenshot as
One time.
As a kind of mode, auditor can add pending video content screenshot according to the violation mark set
Corresponding violation is added to identify, it, can convenient for electronic equipment by being identified in violation of rules and regulations to the video content screenshot addition for including violation information
Whether it is the content screenshot for including violation information with rapid identification video content, memory cost can be saved, then carried
There is the video interception of corresponding mark in violation of rules and regulations.So in such a case, it is possible to obtain from multiple screenshots that video content includes
What is identified includes the screenshot of violation information along video playing progress at first, using the screenshot as first object unit screenshot, into
And it can be using the screenshot time of first object unit screenshot as at the first time.It should be noted that indicating live streaming at the first time
The content of the video practical time for starting violation.
Step S220: obtaining the second object element screenshot, using the screenshot time of the second object element screenshot as the
Two times, the second object element screenshot are to be identified from multiple unit screenshots that video content includes with the second identification method
Out include at first along video playing progress violation information unit screenshot.
Referring to Fig. 6, step S220 may include:
Step S221: the list that the violation label of the second identification method is carried in multiple unit screenshots of video content is obtained
First screenshot.
It should be noted that the second identification method in the present embodiment includes machine recognition.So obtain video content
The unit screenshot that the violation label of the second identification method is carried in multiple unit screenshots may include: to obtain with machine recognition side
What formula was identified from multiple unit screenshots that video content includes includes the unit of violation information along video playing progress at first
Screenshot, using these unit screenshots as the second target screenshot.
As a kind of mode, for the timeliness of verifier identification violation information, live streaming auditing system can allow machine pair
Above-mentioned includes that multiple video interceptions are identified.Optionally, machine can be according to model trained in advance in video content
Whether identified comprising violation information.Optionally, machine can obtain the video content screenshot comprising violation information to identification
It is marked, it should be noted that the label of machine addition can be any form of and can be different from the first identification method
The label identified obtains the video interception for carrying label in violation of rules and regulations.
In a specific application scenarios, when multiple video frame screenshots of the machine to certain section of video content identify
When, machine can be allowed to mark the video interception addition for carrying violation information recognized in violation of rules and regulations, obtain taking for machine recognition
Video interception with violation label, can be using the video interception for carrying label in violation of rules and regulations of these machine recognitions as carrying
The unit screenshot of the violation label of second identification method.
Step S222: along video playing progress in the unit screenshot that the violation for carrying the second identification method is marked
The unit screenshot occurred earliest is as the second object element screenshot, using the screenshot time of the second object element screenshot as
Two times.
As a kind of mode, auditing system is broadcast live can obtain the violation label for carrying the second identification method by machine
Video interception in, the video interception occurred earliest along video playing progress as the second object element screenshot, and then can will
The screenshot time of second object element screenshot is as the second time.
Step S230: the safe period of time between the first time and second time, the safe period of time table are obtained
Levy do not include in the video content between the first time and second time violation information period.
It is understood that violation information can be continuously for each section there are the video content of violation information,
Can be it is discontinuous, if that the video content between screenshots directly by adjacent two comprising violation information is ignored, that
It is likely to cause the discontinuous of normal video content, brings disagreeableness viewing experience to user, therefore, it is necessary to go to obtain
Take the period between the video content comprising discontinuous violation information.
As a kind of mode, auditor can be based on first time obtained above and the second time, when screening first
Between do not include in video content between the second time violation information the safe screenshot period, as first time and
Safe period of time between two times.
Step S240: known based on the safe period of time, from first identification method that same levels mark in violation of rules and regulations
Not Chu the second time for being identified of first time and second identification method, obtain the violation of second identification method
Identify timeliness.
As a kind of mode, auditing system is broadcast live can be by the second time, and at the first time and the sum of safe screenshot period
Difference, as the second identification method violation identify timeliness.Optionally, with reference to foregoing description it is found that violation information packet
A variety of grades are included, then as an implementation, during the violation for calculating the second identification method identifies timeliness, first
Time and the second time need to be from the violation informations of same levels, that is, can disobey based on safe period of time, from same levels
The second time that first time for being identified of the first identification method of rule information and the second identification method are identified obtains the
The violation of two identification methods identifies timeliness.
For example, to Mr. Yu's section video, electronic equipment is getting first object unit in a specific application scenarios
After screenshot and the second object element screenshot, it can first judge that first object unit screenshot and the second object element screenshot are each
From included violation information discipline rating it is whether identical, however, it is determined that first object unit screenshot and the second object element
Respectively in the identical situation of discipline rating of included violation information, the screenshot time of first object unit screenshot is made for screenshot
For that at the first time, using the screenshot time of the second object element screenshot as the second time, and then can guarantee to handle violation video
The fairness of content.
Step S250: carrying out default processing to the video content according to the violation identification timeliness got, to
Reduce the exposure time of the video content.
Wherein, as a kind of mode, auditing system, which is broadcast live, can preset the threshold value for identifying timeliness in violation of rules and regulations, for obtaining
The violation identification timeliness arrived, it can be determined that whether identification timeliness is greater than given threshold in violation of rules and regulations.As a kind of mode, if more than that
Video content can be carried out breaking and broadcast or break processing；Alternatively, if being not more than, then can be to video content
Carry out other processing, such as warning or fine etc..
In a specific application scenarios, if live streaming auditing system has preset the threshold value of machine recognition violation timeliness
It is 30 seconds, is broadcast live in latter 30 minutes of a length of 1 hour video content when certain section of auditing system interception comprising violation information, this
30 minutes video content informations practical time for starting violation is the 45th minute afterwards, and this latter 30 minutes that machine recognition obtains
Video content to start the time of violation be the 45th point include that there is no in the video of violation information in 10 seconds between 55 seconds, this 55 seconds
Hold, then the violation identification timeliness of this rear 30 minutes video content of machine recognition be " 45 points 55 seconds -45 points -10 seconds=45
Second ", since 45 seconds greater than 30 seconds, this rear 30 minutes video content can be carried out breaking and broadcast or break processing.It can keep away
Exempt from excessive video content violation information to expose for a long time in live streaming platform, and then user experience can be promoted.
It is understood that if identification timeliness is less than given threshold in violation of rules and regulations, it, can be by live streaming audit system as a kind of mode
The video content that violation information will be present in system plays out after being filtered.
The present embodiment identifies timeliness by calculating the violation of the second identification method, it can be determined that the recognition capability of machine is made
For a kind of mode, identify that timeliness is bigger in violation of rules and regulations, machine recognition ability is weaker, identifies that timeliness is smaller in violation of rules and regulations, machine recognition ability is got over
By force.Further, identification timeliness is smaller in violation of rules and regulations, it is possible to determine that machine recognition ability is more accurate.
It should be noted that identifying timeliness for the violation that the present embodiment obtains, can be used for according to identification timeliness to straight
The broadcasting situation for broadcasting video is adjusted, for example, if it is detected that can include violation to corresponding when identification timeliness is larger in violation of rules and regulations
The video content of information, which break, broadcasts or breaks processing.Further, timeliness is identified by the violation of computing machine, it can be into one
Which step analysis be the violation information that machine is not covered with when whether machine recognition video content includes violation information there are also
Identification model, and then the ability of violation information can be identified with hoisting machine.
The present embodiment is illustratively illustrated by taking Fig. 7 as an example below:
Referring to Fig. 7, in a specific application scenarios, it is assumed that Fig. 7 is rear 30 minutes views of one section of live video
Frequency content, live streaming auditing system have been carried out carrying out cutting figure with 3 seconds frequencies, obtained comprising multiple by machine to the video content
The video content of screenshot.As a kind of mode, " 1 " in Fig. 7 indicate by auditor by labeling system mark it is above-mentioned after
First is marked with the screenshot of violation reason in 30 minutes video content screenshots, it should be noted that the screenshot time is straight
Broadcast the time T1 that practical violation starts." 2 " indicate to be obtained by machine recognition it is above-mentioned after the in 30 minutes video content screenshots
One is marked with the screenshot of violation reason, and optionally, which indicates that the machine recognition screenshot is doubtful violation.
Optionally, in the presence of the safe screenshot period for not including violation information between screenshot " 1 " and screenshot " 2 ", as one
Kind of mode, can be by obtaining that the first screenshot time comprising violation information subtracts auditor's mark for machine recognition
One includes the screenshot time of violation information, then subtracts the safe screenshot period between screenshot " 1 " and screenshot " 2 ", can obtain
Violation to machine recognition identifies timeliness.The current video content comprising violation information can be carried out according to violation identification timeliness
Processing, such as break and broadcast or break processing etc..
It should be noted that the first identification method of description in the present embodiment and the recognition sequence between the second identification method
Not as restriction, can be adjusted according to the actual situation.
A kind of method handling violation video content provided in this embodiment, by obtaining at the first time；Then is obtained
Two times；The safe period of time at the first time between the second time is obtained again；It is then based on safe period of time, first time and second
Time obtains the violation timeliness of the second identification method；Default place is carried out to video content according to the violation identification timeliness got
Reason, to reduce the exposure time of video content.This method is by obtaining the live streaming violation screenshot time and obtaining machine recognition
The violation screenshot time is broadcast live, by not wrapping between removal live streaming violation screenshot time and the live streaming violation screenshot time of machine recognition
The safe period of time for including violation information obtains the timeliness of machine recognition violation, demonstrates the ability of the timeliness of machine recognition violation, root
Video content break according to violation identification timeliness and broadcasts or break processing, can reduce the exposure time of violation information.
Referring to Fig. 8, for a kind of structural block diagram for the device for handling violation video content provided by the embodiments of the present application, this
Embodiment provides a kind of device 300 for handling violation video content, runs on electronic equipment, and described device 300 includes: first to obtain
Modulus block 310, second obtains module 320, third obtains module 330 and the 4th obtains module 340:
First obtains module 310, and for obtaining at the first time, first time characterization is with the first identification method from video
What is identified in content includes the time of violation information along video playing progress at first, and first identification method is trusty
Identification method.
As a kind of mode, the first acquisition module 310 can be used for obtaining first object unit screenshot, by first object list
The screenshot time of first screenshot is as at the first time, wherein the first object unit screenshot in the present embodiment is with the first identification side
What formula was identified from multiple unit screenshots that video content includes includes the unit of violation information along video playing progress at first
Screenshot.
Optionally, the first acquisition module 310, which can be used for obtaining in multiple unit screenshots of video content, carries first
The unit screenshot of the violation label of identification method；It will carry in the unit screenshot of the violation label of the first identification method along video
The unit screenshot that playback progress occurs earliest makees the screenshot time of first object unit screenshot as first object unit screenshot
For at the first time.
Second obtains module 320, and for obtaining for the second time, second time representation is with the second identification method from video
What is identified in content includes the time of violation information along video playing progress at first, and second identification method is to be detected
Identification method.
As a kind of mode, the second acquisition module 320 can be used for obtaining the second object element screenshot, by the second target list
The screenshot time of first screenshot is as the second time, wherein the second object element screenshot in the present embodiment is with the second identification side
What formula was identified from multiple unit screenshots that video content includes includes the unit of violation information along video playing progress at first
Screenshot.
Optionally, the second acquisition module 320, which can be used for obtaining in multiple unit screenshots of video content, carries second
The unit screenshot of the violation label of identification method；It will carry in the unit screenshot of the violation label of the second identification method along video
The unit screenshot that playback progress occurs earliest makees the screenshot time of the second object element screenshot as the second object element screenshot
For the second time.
In one implementation, the second identification method may include machine recognition mode, and second obtains module 320 can be with
For obtain identified from multiple unit screenshots that video content includes in a manner of machine recognition along video playing progress most
It first include the unit screenshot of violation information；Using the unit screenshot as the second target screenshot.
Third obtains module 330, described for obtaining the safe period of time between the first time and second time
Safe period of time characterize do not include in the video content between the first time and second time violation information period.
As a kind of mode, third, which obtains module 330, can be used for obtaining the safety between the second time at the first time
The screenshot period, it does not include believing in violation of rules and regulations which, which can characterize in the video content between the second time at the first time,
The unit screenshot of breath.
4th obtains module 340, for being based on the safe period of time, the first time and second time, obtains
The violation of second identification method identifies timeliness.
Optionally, label may include that the violation of multiple grades marks in violation of rules and regulations, obtain module as a kind of mode, the 4th
340 can be used for based on safe period of time, the first time identified from the first identification method that same levels mark in violation of rules and regulations
The second time identified with the second identification method, the violation for obtaining the second identification method identify timeliness.
As an implementation, the 4th acquisition module 340 can be used for first time and safety cutting for the second time
The difference for scheming the sum of period, the violation as the second identification method identify timeliness.
Processing module 350, for carrying out default place to the video content according to the violation identification timeliness got
Reason, to reduce the exposure time of the video content.
As a kind of mode, when getting identification timeliness in violation of rules and regulations, when processing module 350 can be used for judging identifying in violation of rules and regulations
Whether effect is greater than given threshold, optionally, broadcasts or break processing if so, can so carry out breaking to video content；If it is not, that
Warning processing or fine processing etc. can be carried out, to video content so as to reduce the video content for including violation information
Exposure time.
A kind of device handling violation video content provided in this embodiment, by obtaining at the first time；Then is obtained
Two times；The safe period of time at the first time between the second time is obtained again；It is then based on safe period of time, first time and second
Time, the violation for obtaining the second identification method identify timeliness；Video content is carried out according to the violation identification timeliness got pre-
If processing, to reduce the exposure time of video content.By obtaining the live streaming violation screenshot time and obtaining the straight of machine recognition
The violation screenshot time is broadcast, by not including between removal live streaming violation screenshot time and the live streaming violation screenshot time of machine recognition
The safe period of time of violation information obtains the timeliness of machine recognition violation, demonstrates the ability of the timeliness of machine recognition violation, then
Person break to video content and broadcasts or break processing, can reduce the exposure time of violation information according to violation identification timeliness.
It is apparent to those skilled in the art that for convenience and simplicity of description, foregoing description device and
The specific work process of module, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In several embodiments provided herein, the mutual coupling of shown or discussed module or direct coupling
It closes or communication connection can be through some interfaces, the indirect coupling or communication connection of device or module can be electrical property, mechanical
Or other forms.
It, can also be in addition, can integrate in a processing module in each functional module in each embodiment of the application
It is that modules physically exist alone, can also be integrated in two or more modules in a module.Above-mentioned integrated mould
Block both can take the form of hardware realization, can also be realized in the form of software function module.
Referring to Fig. 9, the method and device based on above-mentioned processing violation video content, the embodiment of the present application also provides
A kind of electronic equipment 12 for the method that can execute aforementioned processing violation video content.Electronic equipment 12 include memory 122 with
And one or more (one is only shown in the figure) processor 124 to intercouple, lead between memory 122 and processor 124
Believe connection.The program that can execute content in previous embodiment is stored in memory 122, and processor 124 can be held
The program stored in line storage 122.
Wherein, processor 124 may include one or more processing core.Processor 124 utilizes various interfaces and route
The various pieces in entire electronic equipment 12 are connected, by running or executing the instruction being stored in memory 122, program, generation
Code collection or instruction set, and the data being stored in memory 122 are called, execute the various functions and processing of electronic equipment 12
Data.Optionally, processor 124 can use Digital Signal Processing (Digital Signal Processing, DSP), show
Field programmable gate array (Field-Programmable Gate Array, FPGA), programmable logic array
At least one of (Programmable Logic Array, PLA) example, in hardware is realized.Processor 124 can integrating central
Processor (Central Processing Unit, CPU), image processor (Graphics Processing Unit, GPU)
With the combination of one or more of modem etc..Wherein, the main processing operation system of CPU, user interface and apply journey
Sequence etc.；GPU is for being responsible for the rendering and drafting of display content；Modem is for handling wireless communication.It is understood that
Above-mentioned modem can not also be integrated into processor 124, be realized separately through one piece of communication chip.
Memory 122 may include random access memory (Random Access Memory, RAM), also may include read-only
Memory (Read-Only Memory).Memory 122 can be used for store instruction, program, code, code set or instruction set.It deposits
Reservoir 122 may include storing program area and storage data area, wherein the finger that storing program area can store for realizing operating system
Enable, for realizing at least one function instruction (such as touch function, sound-playing function, image player function etc.), be used for
Realize the instruction etc. of foregoing individual embodiments.Storage data area can also store the data that electronic equipment 12 is created in use
(such as phone directory, audio, video data, chat record data) etc..
Referring to FIG. 10, it illustrates a kind of structural frames of computer readable storage medium provided by the embodiments of the present application
Figure.Program code is stored in the computer readable storage medium 400, said program code can call execution above-mentioned by processor
Method described in embodiment of the method.
Computer readable storage medium 400 can be such as flash memory, EEPROM (electrically erasable programmable read-only memory),
The electronic memory of EPROM, hard disk or ROM etc.Optionally, computer readable storage medium 400 includes non-transient meter
Calculation machine readable medium (non-transitory computer-readable storage medium).Computer-readable storage
Medium 400 has the memory space for the program code 410 for executing any method and step in the above method.These program codes can
With from reading or be written in one or more computer program product in this one or more computer program product.
Program code 410 can for example be compressed in a suitable form.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is contained at least one embodiment or example of the application.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
Finally, it should be noted that above embodiments are only to illustrate the technical solution of the application, rather than its limitations；Although
The application is described in detail with reference to the foregoing embodiments, those skilled in the art are when understanding: it still can be with
It modifies the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；And
These are modified or replaceed, do not drive corresponding technical solution essence be detached from each embodiment technical solution of the application spirit and
Range.
Claims (11)
1. a kind of method for handling violation video content, which is characterized in that the described method includes:
Obtain at the first time, first time characterization with the first identification method identify from video content along video playing
Progress includes the time of violation information at first, and first identification method is identification method trusty；
Obtained for the second time, second time representation with the second identification method identify from video content along video playing
Progress includes the time of violation information at first, and second identification method is identification method to be detected；
The safe period of time between the first time and second time is obtained, the safe period of time characterizes the first time
It does not include the period of violation information in video content between second time；
Based on the safe period of time, the first time and second time, the violation for obtaining second identification method is known
Other timeliness；
Default processing is carried out to the video content according to the violation identification timeliness got, to reduce in the video
The exposure time of appearance.
2. the method as described in claim 1, which is characterized in that the violation identification timeliness that the basis is got is to described
Video content carries out presetting the step of handling
Judge whether the violation identification timeliness is greater than given threshold；
Processing is broadcast or broken if so, break to the video content.
3. the method as described in claim 1, which is characterized in that the step of acquisition first time includes:
First object unit screenshot is obtained, using the screenshot time of the first object unit screenshot as first time, described the
One object element screenshot is to be broadcast with what the first identification method was identified from multiple unit screenshots that video content includes along video
Degree of putting into includes the unit screenshot of violation information at first；
The step of second time of the acquisition includes:
The second object element screenshot is obtained, using the screenshot time of the second object element screenshot as the second time, described
Two object element screenshots are to be broadcast with what the second identification method was identified from multiple unit screenshots that video content includes along video
Degree of putting into includes the unit screenshot of violation information at first.
4. method as claimed in claim 3, which is characterized in that the acquisition first object unit screenshot, by first mesh
Mark unit screenshot the screenshot time include: as the step of first time
Obtain the unit screenshot that the violation label of the first identification method is carried in multiple unit screenshots of video content；
The list that will occur earliest in the unit screenshot of the violation label for carrying the first identification method along video playing progress
First screenshot is as first object unit screenshot, using the screenshot time of the first object unit screenshot as at the first time；
The second object element screenshot of the acquisition, using the screenshot time of the second object element screenshot as the step of the second time
Suddenly include:
Obtain the unit screenshot that the violation label of the second identification method is carried in multiple unit screenshots of video content；
The list that will occur earliest in the unit screenshot of the violation label for carrying the second identification method along video playing progress
First screenshot is as the second object element screenshot, using the screenshot time of the second object element screenshot as the second time.
5. method as claimed in claim 4, which is characterized in that the violation label includes that the violation of multiple grades marks, institute
State the violation identification that second identification method is obtained based on the safe period of time, the first time and second time
The step of timeliness includes:
Based on the safe period of time, the first time identified from first identification method that same levels mark in violation of rules and regulations
The second time identified with second identification method, the violation for obtaining second identification method identify timeliness.
6. method as claimed in claim 3, which is characterized in that second identification method includes machine recognition mode, is obtained
Second object element screenshot, the screenshot time using the second object element screenshot includes: as the step of the second time
Obtain identified from multiple unit screenshots that video content includes in a manner of machine recognition along video playing progress most
It first include the unit screenshot of violation information；
Using the unit screenshot as the second target screenshot.
7. method as claimed in claim 3, which is characterized in that between the acquisition first time and second time
Safe period of time the step of include:
The safe screenshot period between the first time and second time is obtained, described in the safe screenshot period characterization
It does not at the first time include the unit screenshot of violation information in the video content between second time.
8. the method for claim 7, which is characterized in that described to be based on the safe screenshot period, the first time
With second time, obtain second method violation identification timeliness the step of include:
By second time, difference with the sum of the first time and the safe screenshot period, as the second identification side
The violation of formula identifies timeliness.
9. a kind of device for handling violation video content characterized by comprising
First obtains module, and for obtaining at the first time, the first time is characterized with the first identification method from video content
What is identified includes the time of violation information along video playing progress at first, and first identification method is identification side trusty
Formula；
Second obtain module, for obtain the second time, second time representation with the second identification method from video content
What is identified includes the time of violation information along video playing progress at first, and second identification method is identification side to be detected
Formula；
Third obtains module, for obtaining the safe period of time between the first time and second time, when described safe
Segment table levy do not include in the video content between the first time and second time violation information period；
4th obtains module, for being based on the safe period of time, the first time and second time, obtains described second
The violation of identification method identifies timeliness；
Processing module, for carrying out default processing to the video content according to the violation identification timeliness got, to
Reduce the exposure time of the video content.
10. a kind of electronic equipment, which is characterized in that including memory；
One or more processors；
One or more programs are stored in the memory and are configured as being executed by one or more of processors, institute
It states one or more programs and is configured to carry out any method of claim 1-8.
11. a kind of computer readable storage medium, which is characterized in that be stored with program generation in the computer readable storage medium
Code, wherein perform claim requires any method of 1-8 when said program code is run by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811626292.4A CN109831696B (en) | 2018-12-28 | 2018-12-28 | Method and device for processing illegal video content, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811626292.4A CN109831696B (en) | 2018-12-28 | 2018-12-28 | Method and device for processing illegal video content, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109831696A true CN109831696A (en) | 2019-05-31 |
| CN109831696B CN109831696B (en) | 2021-03-12 |
Family
ID=66860598
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811626292.4A Active CN109831696B (en) | 2018-12-28 | 2018-12-28 | Method and device for processing illegal video content, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109831696B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111866605A (en) * | 2020-07-09 | 2020-10-30 | 北京齐尔布莱特科技有限公司 | Video auditing method and server |
| CN112911321A (en) * | 2021-01-22 | 2021-06-04 | 上海七牛信息技术有限公司 | Real Time Clock (RTC) -based audio and video live broadcast control method and system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103914530A (en) * | 2014-03-31 | 2014-07-09 | 北京中科模识科技有限公司 | Method and system for monitoring rule-violating advertisements in broadcasting and TV programs |
| CN106454492A (en) * | 2016-10-12 | 2017-02-22 | 武汉斗鱼网络科技有限公司 | Live pornographic content audit system and method based on delayed transmission |
| US20170142181A1 (en) * | 2015-11-16 | 2017-05-18 | Le Holdings (Beijing) Co., Ltd. | Webcasting method, device and system |
| CN107609598A (en) * | 2017-09-27 | 2018-01-19 | 武汉斗鱼网络科技有限公司 | Image authentication model training method, device and readable storage medium storing program for executing |
| CN108154134A (en) * | 2018-01-11 | 2018-06-12 | 天格科技（杭州）有限公司 | Internet live streaming pornographic image detection method based on depth convolutional neural networks |
| CN108419091A (en) * | 2018-03-02 | 2018-08-17 | 北京未来媒体科技股份有限公司 | A kind of verifying video content method and device based on machine learning |
| US20180247305A1 (en) * | 2007-12-19 | 2018-08-30 | At&T Intellectual Property I, L.P. | Systems and methods to identify target video content |
| CN108833962A (en) * | 2018-05-25 | 2018-11-16 | 咪咕音乐有限公司 | Display information processing method and device and storage medium |
| CN108900910A (en) * | 2018-07-16 | 2018-11-27 | 上海艾策通讯科技股份有限公司 | Monitor the method and system of IPTV service content legality |
-
2018
- 2018-12-28 CN CN201811626292.4A patent/CN109831696B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180247305A1 (en) * | 2007-12-19 | 2018-08-30 | At&T Intellectual Property I, L.P. | Systems and methods to identify target video content |
| CN103914530A (en) * | 2014-03-31 | 2014-07-09 | 北京中科模识科技有限公司 | Method and system for monitoring rule-violating advertisements in broadcasting and TV programs |
| US20170142181A1 (en) * | 2015-11-16 | 2017-05-18 | Le Holdings (Beijing) Co., Ltd. | Webcasting method, device and system |
| CN106454492A (en) * | 2016-10-12 | 2017-02-22 | 武汉斗鱼网络科技有限公司 | Live pornographic content audit system and method based on delayed transmission |
| CN107609598A (en) * | 2017-09-27 | 2018-01-19 | 武汉斗鱼网络科技有限公司 | Image authentication model training method, device and readable storage medium storing program for executing |
| CN108154134A (en) * | 2018-01-11 | 2018-06-12 | 天格科技（杭州）有限公司 | Internet live streaming pornographic image detection method based on depth convolutional neural networks |
| CN108419091A (en) * | 2018-03-02 | 2018-08-17 | 北京未来媒体科技股份有限公司 | A kind of verifying video content method and device based on machine learning |
| CN108833962A (en) * | 2018-05-25 | 2018-11-16 | 咪咕音乐有限公司 | Display information processing method and device and storage medium |
| CN108900910A (en) * | 2018-07-16 | 2018-11-27 | 上海艾策通讯科技股份有限公司 | Monitor the method and system of IPTV service content legality |
Non-Patent Citations (1)
| Title |
|---|
| 韩子寅: "《互联网视频监管系统影标识别算法的研究》", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111866605A (en) * | 2020-07-09 | 2020-10-30 | 北京齐尔布莱特科技有限公司 | Video auditing method and server |
| CN111866605B (en) * | 2020-07-09 | 2022-10-18 | 北京齐尔布莱特科技有限公司 | Video auditing method and server |
| CN112911321A (en) * | 2021-01-22 | 2021-06-04 | 上海七牛信息技术有限公司 | Real Time Clock (RTC) -based audio and video live broadcast control method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109831696B (en) | 2021-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8583509B1 (en) | Method of and system for fact checking with a camera device | |
| CN111640436B (en) | Method for providing dynamic customer portraits of conversation objects to agents | |
| CN110689438A (en) | Enterprise financial risk scoring method and device, computer equipment and storage medium | |
| CN113383362B (en) | User identification method and related product | |
| CN105869008A (en) | Targeted delivery method and device of advertisement | |
| CN107807941A (en) | Information processing method and device | |
| CN108399565A (en) | Financial product recommendation apparatus, method and computer readable storage medium | |
| CN109803152A (en) | Violation checking method, device, electronic equipment and storage medium | |
| CN109766472A (en) | Signal auditing method, device, electronic equipment and storage medium | |
| CN111008887B (en) | Skill service bidding system | |
| CN111476653A (en) | Risk information identification, determination and model training method and device | |
| US20200211061A1 (en) | Method and system for targeted advertising in an over-the-top (ott) ecosystem | |
| CN109831696A (en) | Handle method, apparatus, electronic equipment and the storage medium of violation video content | |
| CN112307464A (en) | Fraud identification method and device and electronic equipment | |
| CN113129012A (en) | Payment data processing method, device, equipment and system | |
| CN111695357A (en) | Text labeling method and related product | |
| CN114119030A (en) | Fraud prevention method and device, electronic equipment and storage medium | |
| KR101972660B1 (en) | System and Method for Checking Fact | |
| Angus et al. | The Australian Ad Observatory technical and data report | |
| CN113010785A (en) | User recommendation method and device | |
| King et al. | Lying about lying on social media: a case study of the 2019 Canadian Elections | |
| CN116362534A (en) | Emergency management method and system for violations and risks of online customer service contents in railway field | |
| CN114239561A (en) | Supply relation obtaining method and device, storage medium and electronic equipment | |
| CN111105263B (en) | User identification method, device, electronic equipment and storage medium | |
| CN113449506A (en) | Data detection method, device and equipment and readable storage medium |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20190531 Assignee: GUANGZHOU CUBESILI INFORMATION TECHNOLOGY Co.,Ltd. Assignor: GUANGZHOU HUADUO NETWORK TECHNOLOGY Co.,Ltd. Contract record no.: X2021440000032 Denomination of invention: Method, device, electronic device and storage medium for processing illegal video content License type: Common License Record date: 20210125 |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |