Video content moderation is an offline task that intelligently moderates video content with the aid of AI. The task execution results include confidence score, moderation suggestion, and suspected video segments. According to the suggestion, you can decide whether to allow a video to be published, effectively avoiding potential legal risks and brand image damage caused by illegal videos.
VOD can intelligently detect porn, terrorism, and politically sensitive information in video image, speech (ASR), and text (OCR).
Object | Operation | Description |
---|---|---|
Video image (figures and objects) | ||
Porn information detection |
Performs porn information detection on video image, including:
|
|
Terrorism information detection |
Performs terrorism information detection on video image, including:
|
|
Politically sensitive information detection |
Performs politically sensitive information detection on video image, including:
|
|
ASR phrase (phrase in speech) | Porn information detection | Performs porn information detection on phrases in speech to identify suspect keywords |
Politically sensitive information detection | Performs politically sensitive information detection on phrases in speech to identify suspect keywords | |
OCR text (text in video image) | Porn information detection | Performs porn information detection on text in video image to identify suspect keywords |
Politically sensitive information detection | Performs politically sensitive information detection on text in video image to identify suspect keywords |
Field Name | Type | Description |
---|---|---|
confidence | Float | Confidence score (0–100). The higher the score, the greater the suspicion |
suggestion | String | There are three types of moderation suggestions:
|
segments | Array | Suspected video segments, helping locate specific segments in the video that are suspected of violations |
A video moderation template represents a set of video moderation parameters. You can use a template to specify which of the following video moderation tasks you want VOD to perform:
VOD provides preset video moderation templates for common parameter combinations. You can also use a server API to create and manage custom templates.
You can initiate a video moderation task by calling a server API, via the console, or by specifying the task when uploading videos. For details, see Task Initiation.
Below are instructions for initiating video content moderation tasks in these ways:
AiContentReviewTask
parameter in the request.AiContentReviewTask
parameter in the request.procedure
to the name of the task flow when calling ApplyUpload.procedure
to the name of the task flow in the signature for upload from client.After initiating a video moderation task, you can wait for result notification asynchronously or perform task query synchronously to get the task execution result. Below is an example of getting the result notification in normal callback mode after the video moderation task is initiated (the fields with null value are omitted):
{
"EventType":"ProcedureStateChanged",
"ProcedureStateChangeEvent":{
"TaskId":"1256768367-Procedure-2e1af2456351812be963e309cc133403t0",
"Status":"FINISH",
"FileId":"5285890784246869930",
"FileName":"Animal World",
"FileUrl":"http://1256768367.vod2.myqcloud.com/xxx/xxx/AtUCmy6gmIYA.mp4",
"MetaData":{
"AudioDuration":60,
"AudioStreamSet":[
{
"Bitrate":383854,
"Codec":"aac",
"SamplingRate":48000
}
],
"Bitrate":1021028,
"Container":"mov,mp4,m4a,3gp,3g2,mj2",
"Duration":60,
"Height":480,
"Rotate":0,
"Size":7700180,
"VideoDuration":60,
"VideoStreamSet":[
{
"Bitrate":637174,
"Codec":"h264",
"Fps":23,
"Height":480,
"Width":640
}
],
"Width":640
},
"AiContentReviewResultSet":[
{
"Type":"Porn",
"PornTask":{
"Status":"SUCCESS",
"ErrCode":0,
"Message":"",
"Input":{
"Definition":10
},
"Output":{
"Confidence":98,
"Suggestion":"block",
"Label":"sexy",
"SegmentSet":[
{
"StartTimeOffset":9.5,
"EndTimeOffset":14,
"Confidence":98,
"Suggestion":"block",
"Label":"sexy",
"Url":"http://xxx.vod2.myqcluod.com/xxx/xxx/xx1.jpg",
"PicUrlExpireTimeStamp":1530005146
},
{
"StartTimeOffset":16.5,
"EndTimeOffset":18,
"Confidence":80,
"Suggestion":"review",
"Label":"sexy",
"Url":"http://xxx.vod2.myqcluod.com/xxx/xxx/xx2.jpg",
"PicUrlExpireTimeStamp":1530005146
},
{
"StartTimeOffset":41,
"EndTimeOffset":49,
"Confidence":97,
"Suggestion":"block",
"Label":"sexy",
"Url":"http://xxx.vod2.myqcluod.com/xxx/xxx/xx3.jpg",
"PicUrlExpireTimeStamp":1530005146
}
]
}
}
},
{
"Type":"Terrorism",
"TerrorismTask":{
"Status":"SUCCESS",
"ErrCode":0,
"Message":"",
"Input":{
"Definition":10
},
"Output":{
"Confidence":0,
"Suggestion":"pass",
"SegmentSet":[
]
}
}
},
{
"Type":"Political",
"PoliticalTask":{
"Status":"SUCCESS",
"ErrCode":0,
"Message":"",
"Input":{
"Definition":10
},
"Output":{
"Confidence":0,
"Suggestion":"pass",
"SegmentSet":[
]
}
}
}
],
"TasksPriority":0,
"TasksNotifyMode":""
}
}
In the callback result, ProcedureStateChangeEvent.AiContentReviewResultSet
contains three types of moderation results, namely Porn
, Terrorism
, and Political
, which represent the detection of porn, terrorism, and politically sensitive information in video images, respectively.
Type
of Porn
shows that Output.Suggestion
is block
, that is, the possibility of porn information presence is high and blocking is recommended, the confidence of porn information is 98, and the reason is sexy
(sexiness).Output.SegmentSet
in Type
of Porn
lists three video segments suspected of containing porn information. The start time and end time of each segment are marked by StartTimeOffset
and EndTimeOffset
, respectively.Type
of Terrorism
and Political
shows that the video is not suspected of containing terrorism and politically sensitive information.
Was this page helpful?