檢查 Azure AI Video Indexer 輸出
警告
在過去的一年裡,Azure AI 影片索引器(VI)宣佈因淘汰而取消對 Azure 媒體服務 (AMS) 的相依性。 已宣佈功能調整 和 變更 ,並提供 移轉指南 。
完成移轉的最後期限是 2024 年 6 月 30 日。 VI 已延長更新/移轉期限,讓您可以更新VI帳戶,並選擇加入AMS VI資產移轉至2024年8月31日。
不過,在 6 月 30 日之後,如果您尚未更新 VI 帳戶,您將無法編製新影片的索引,也無法播放尚未移轉的任何影片。 如果您在 6 月 30 日之後更新您的帳戶,您可以立即繼續編製索引,但 您將無法在帳戶更新之前播放影片,直到透過 AMS VI 移轉進行移轉。
編製影片索引時,Azure AI Video Indexer 會產生 JSON 內容,其中包含所指定影片深入解析的詳細資料。 這些深入解析包括文字記錄、光學字元辨識元素 (OCR)、臉部、主題以及類似的詳細資料。 每個深入解析類型都包含時間範圍的執行個體,其會顯示影片中的深入解析出現的時間。
如需詳細資訊,請參閱 Azure AI Video Indexer 深入解析。
深入解析的根項目
名稱 | 描述 |
---|---|
accountId |
播放清單的 VI 帳戶識別碼。 |
id |
播放清單的識別碼。 |
name |
播放清單的名稱。 |
description |
播放清單的描述。 |
userName |
建立播放清單的使用者名稱。 |
created |
播放清單的建立時間。 |
privacyMode |
播放清單的隱私權模式 (Private 或 Public )。 |
state |
播放清單的狀態 (Uploaded Processing Processed Failed 或 Quarantined )。 |
isOwned |
指出目前使用者是否已建立播放清單。 |
isEditable |
指出目前的使用者是否有權編輯播放清單。 |
isBase |
指出播放清單是基礎播放清單 (影片) 還是以其他影片組成的播放清單 (衍生)。 |
durationInSeconds |
播放清單的持續時間總計。 |
summarizedInsights |
產生的 JSON 輸出包含 Insights 與 SummarizedInsights 元素。 建議您使用 Insights ,而非 SummarizedInsights (為提供回溯相容性)。 |
videos |
建構播放清單的影片清單。 如果此播放清單是由其他影片 (衍生) 的時間範圍建構而成,則此清單中的影片只會包含來自所含時間範圍的資料。 |
{
...
"accountId": "00000000-0000-0000-0000-000000000000",
"id": "abc3454321",
"name": "My first video",
"description": "I am trying VI",
"userName": "Some name",
"created": "2018/2/2 18:00:00.000",
"privacyMode": "Private",
"state": "Processed",
"isOwned": true,
"isEditable": false,
"isBase": false,
"durationInSeconds": 120,
"summarizedInsights" : null,
"videos": [{ . . . }]
}
提示
產生的 JSON 輸出包含 Insights
與 SummarizedInsights
元素。 強烈建議您使用 Insights
,而非 SummarizedInsights
(為了提供回溯相容性)。
深入解析的摘要
本節說明深入解析的摘要。
屬性 | 描述 |
---|---|
name |
影片的名稱。 例如: Azure Monitor 。 |
id |
影片的識別碼。 例如: 63c6d532ff 。 |
privacyMode |
您的明細可以具有下列其中一種模式:您帳戶中的每個人和具有影片連結的任何人都可看到 Public 影片。 您帳戶中的每個人都可看到 Private 影片。 |
duration |
深入解析發生的時間,以秒為單位。 |
thumbnailVideoId |
從中取得縮圖的影片識別碼。 |
thumbnailId |
影片的縮圖識別碼。 若要取得實際縮圖,請呼叫 Get-Thumbnail,並將 thumbnailVideoId 和 thumbnailId 傳遞至其中。 |
faces |
包含零個或多個臉部。 如需詳細資訊,請參閱 faces。 |
keywords |
包含零個或多個關鍵字。 如需詳細資訊,請參閱 keywords。 |
sentiments |
包含零個或多個情緒。 如需詳細資訊,請參閱 sentiments。 |
audioEffects |
包含零個或多個音訊效果。 如需詳細資訊,請參閱 audioEffects。 |
labels |
包含零個或多個標籤。 如需詳細資訊,請參閱 labels。 |
brands |
包含零個或多個品牌。 如需詳細資訊,請參閱 brands。 |
statistics |
如需詳細資訊,請參閱統計資料。 |
emotions |
包含零個或多個表情。 如需詳細資訊,請參閱 emotions。 |
topics |
包含零個或多個主題。 如需詳細資訊,請參閱 topics。 |
videos
名稱 | 描述 |
---|---|
accountId |
影片的 VI 帳戶識別碼。 |
id |
影片的識別碼。 |
name |
影片的名稱。 |
state |
影片的狀態 (Uploaded Processing Processed Failed 或 Quarantined )。 |
processingProgress |
處理期間的進度。 例如: 20% 。 |
failureCode |
影片無法處理時的失敗碼。 例如: UnsupportedFileType 。 |
failureMessage |
影片無法處理時的失敗訊息。 |
externalId |
影片的外部識別碼 (如果使用者指定一個)。 |
externalUrl |
影片的外部 URL (如果使用者指定一個)。 |
metadata |
影片的外部中繼資料 (如果使用者指定一個)。 |
isAdult |
指出是否已手動檢閱影片並識別為成人影片。 |
insights |
深入解析物件。 如需詳細資訊,請參閱深入解析。 |
thumbnailId |
影片的縮圖識別碼。 若要取得實際縮圖,請呼叫 Get-Thumbnail,並將影片識別碼和縮圖識別碼傳遞至其中。 |
publishedUrl |
串流影片的 URL。 |
publishedUrlProxy |
在 Apple 裝置上串流影片的 URL。 |
viewToken |
用來串流影片的短期檢視權杖。 |
sourceLanguage |
影片的來源語言。 |
language |
影片的實際語言 (翻譯)。 |
indexingPreset |
用來編製影片索引的預設值。 |
streamingPreset |
用來發佈影片的預設值。 |
linguisticModelId |
用來轉譯影片的文字記錄自訂 (CRIS) 模型。 |
statistics |
如需詳細資訊,請參閱統計資料。 |
{
"videos": [{
"accountId": "2cbbed36-1972-4506-9bc7-55367912df2d",
"id": "142a356aa6",
"state": "Processed",
"privacyMode": "Private",
"processingProgress": "100%",
"failureCode": "General",
"failureMessage": "",
"externalId": null,
"externalUrl": null,
"metadata": null,
"insights": {. . . },
"thumbnailId": "89d7192c-1dab-4377-9872-473eac723845",
"publishedUrl": "https://videvmediaservices.streaming.mediaservices.windows.net:443/d88a652d-334b-4a66-a294-3826402100cd/Xamarine.ism/manifest",
"publishedProxyUrl": null,
"viewToken": "Bearer=<token>",
"sourceLanguage": "En-US",
"language": "En-US",
"indexingPreset": "Default",
"linguisticModelId": "00000000-0000-0000-0000-000000000000"
}],
}
insights
每個深入解析 (例如,文字記錄行、臉部或品牌) 都包含唯一元素清單 (例如,face1
、face2
、face3
)。 每個元素都有自己的中繼資料及其執行個體的清單,這些執行個體是具有額外中繼資料的時間範圍。
臉部可能具有識別碼、名稱、縮圖、其他中繼資料,以及其時態性執行個體的清單 (例如,00:00:05 – 00:00:10
、00:01:00 - 00:02:30
和 00:41:21 – 00:41:49
)。 每個時態性執行個體都可以有額外的中繼資料。 例如,中繼資料可以包含臉部的矩形座標 (20,230,60,60
)。
版本 | 程式碼版本 |
---|---|
sourceLanguage |
影片的來源語言 (假設一個主要語言),格式為 BCP-47 字串。 |
language |
深入解析語言 (從來源語言翻譯),格式為 BCP-47 字串。 |
transcript |
transcript 深入解析。 |
ocr |
OCR 深入解析。 |
keywords |
keywords 深入解析。 |
transcripts |
可能含一或多個文字記錄。 |
faces |
faces 深入解析。 |
labels |
labels 深入解析。 |
shots |
shots 深入解析。 |
brands |
brands 深入解析。 |
audioEffects |
audioEffects 深入解析。 |
sentiments |
sentiments 深入解析。 |
visualContentModeration |
visualContentModeration 深入解析。 |
textualContentModeration |
textualContentModeration 深入解析。 |
emotions |
emotions 深入解析。 |
topics |
topics 深入解析。 |
speakers |
speakers 深入解析。 |
範例:
{
"version": "0.9.0.0",
"sourceLanguage": "en-US",
"language": "es-ES",
"transcript": ...,
"ocr": ...,
"keywords": ...,
"faces": ...,
"labels": ...,
"shots": ...,
"brands": ...,
"audioEffects": ...,
"sentiments": ...,
"visualContentModeration": ...,
"textualContentModeration": ...
}
轉錄
名稱 | 描述 |
---|---|
id |
行識別碼。 |
text |
文字記錄本身。 |
confidence |
文字記錄精確度的信賴等級。 |
speakerId |
說話者的識別碼。 |
language |
文字記錄語言。 旨在支援文字記錄,其中每一行可有不同的語言。 |
instances |
這一行曾出現的時間範圍清單。 如果執行個體在文字記錄中,則其只有一個執行個體。 |
範例:
"transcript":[
{
"id":1,
"text":"Well, good morning everyone and welcome to",
"confidence":0.8839,
"speakerId":1,
"language":"en-US",
"instances":[
{
"adjustedStart":"0:00:10.21",
"adjustedEnd":"0:00:12.81",
"start":"0:00:10.21",
"end":"0:00:12.81"
}
]
},
{
"id":2,
"text":"ignite 2016. Your mission at Microsoft is to empower every",
"confidence":0.8944,
"speakerId":2,
"language":"en-US",
"instances":[
{
"adjustedStart":"0:00:12.81",
"adjustedEnd":"0:00:17.03",
"start":"0:00:12.81",
"end":"0:00:17.03"
}
]
}
ocr
名稱 | 描述 |
---|---|
id |
OCR 的行識別碼。 |
text |
OCR 的文字。 |
confidence |
辨識信賴。 |
language |
OCR 的語言。 |
instances |
此 OCR 曾出現的時間範圍清單。 (相同的 OCR 可以出現多次。) |
height |
OCR 矩形的高度。 |
top |
頂端位置,以像素為單位。 |
left |
左側位置,以像素為單位。 |
width |
OCR 矩形的寬度。 |
angle |
OCR 矩形的角度,從 -180 到 180 。 0 值表示由左至右水平方向。 90 值表示由上至下垂直方向。 180 值表示由右至左水平方向。 -90 值表示由下至上垂直方向。 30 值表示從左上方到右下方。 |
"ocr": [
{
"id": 0,
"text": "LIVE FROM NEW YORK",
"confidence": 675.971,
"height": 35,
"language": "en-US",
"left": 31,
"top": 97,
"width": 400,
"angle": 30,
"instances": [
{
"start": "00:00:26",
"end": "00:00:52"
}
]
}
],
關鍵字
名稱 | 描述 |
---|---|
id |
關鍵字的識別碼。 |
text |
關鍵字的文字。 |
confidence |
關鍵字的辨識信賴度。 |
language |
關鍵字語言 (轉譯時)。 |
instances |
此關鍵字曾出現的時間範圍清單。 (一個關鍵字可以出現多次。) |
{
id: 0,
text: "technology",
confidence: 1,
language: "en-US",
instances: [{
adjustedStart: "0:05:15.782",
adjustedEnd: "0:05:16.249",
start: "0:05:15.782",
end: "0:05:16.249"
},
{
adjustedStart: "0:04:54.761",
adjustedEnd: "0:04:55.228",
start: "0:04:54.761",
end: "0:04:55.228"
}]
}
臉部
如果有臉部,Azure AI Video Indexer 會在所有影片畫面格上使用臉部 API 來偵測臉部和名人。
名稱 | 描述 |
---|---|
id |
臉部的識別碼。 |
name |
臉部的名稱。 其可以是 Unknown #0 、已識別的名人或客戶培訓人員。 |
confidence |
臉部識別的信賴等級。 |
description |
名人的描述。 |
thumbnailId |
臉部的縮圖識別碼。 |
knownPersonId |
如果是已知人員,則為內部識別碼。 |
referenceId |
如果是 Bing 名人,則為 Bing 識別碼。 |
referenceType |
目前只有 Bing。 |
title |
如果是名人,則為人員的頭銜。 例如: Microsoft's CEO 。 |
imageUrl |
如果是名人,則為影像 URL。 |
instances |
其中臉部曾出現在時間範圍的執行個體。 每個執行個體也有一個 thumbnailsIds 值。 |
"faces": [{
"id": 2002,
"name": "Xam 007",
"confidence": 0.93844,
"description": null,
"thumbnailId": "00000000-aee4-4be2-a4d5-d01817c07955",
"knownPersonId": "8340004b-5cf5-4611-9cc4-3b13cca10634",
"referenceId": null,
"title": null,
"imageUrl": null,
"instances": [{
"thumbnailsIds": ["00000000-9f68-4bb2-ab27-3b4d9f2d998e",
"cef03f24-b0c7-4145-94d4-a84f81bb588c"],
"adjustedStart": "00:00:07.2400000",
"adjustedEnd": "00:00:45.6780000",
"start": "00:00:07.2400000",
"end": "00:00:45.6780000"
},
{
"thumbnailsIds": ["00000000-51e5-4260-91a5-890fa05c68b0"],
"adjustedStart": "00:10:23.9570000",
"adjustedEnd": "00:10:39.2390000",
"start": "00:10:23.9570000",
"end": "00:10:39.2390000"
}]
}]
標籤
名稱 | 描述 |
---|---|
id |
標籤的識別碼。 |
name |
標籤的名稱。 例如:Computer 或 TV 。 |
language |
標籤名稱的語言 (已翻譯時),格式為 BCP-47 字串。 |
instances |
此標籤曾出現的時間範圍清單。 (一個標籤可以出現多次。)每個執行個體都有一個信賴度欄位。 |
"labels": [
{
"id": 0,
"name": "person",
"language": "en-US",
"instances": [
{
"confidence": 1.0,
"start": "00: 00: 00.0000000",
"end": "00: 00: 25.6000000"
},
{
"confidence": 1.0,
"start": "00: 01: 33.8670000",
"end": "00: 01: 39.2000000"
}
]
},
{
"name": "indoor",
"language": "en-US",
"id": 1,
"instances": [
{
"confidence": 1.0,
"start": "00: 00: 06.4000000",
"end": "00: 00: 07.4670000"
},
{
"confidence": 1.0,
"start": "00: 00: 09.6000000",
"end": "00: 00: 10.6670000"
},
{
"confidence": 1.0,
"start": "00: 00: 11.7330000",
"end": "00: 00: 20.2670000"
},
{
"confidence": 1.0,
"start": "00: 00: 21.3330000",
"end": "00: 00: 25.6000000"
}
]
}
]
場景
名稱 | 描述 |
---|---|
id |
場景的識別碼。 |
instances |
此場景的時間範圍清單。 (一個場景只能有一個執行個體。) |
"scenes":[
{
"id":0,
"instances":[
{
"start":"0:00:00",
"end":"0:00:06.34",
"duration":"0:00:06.34"
}
]
},
{
"id":1,
"instances":[
{
"start":"0:00:06.34",
"end":"0:00:47.047",
"duration":"0:00:40.707"
}
]
},
]
擷取畫面
名稱 | 描述 |
---|---|
id |
擷取畫面的識別碼。 |
keyFrames |
擷取畫面內的主要畫面格清單。 每個主要畫面格都有一個識別碼和執行個體時間範圍清單。 每個主要畫面格執行個體都有一個 thumbnailId 欄位,其中保留主要畫面格的縮圖識別碼。 |
instances |
此擷取畫面的時間範圍清單。 (一個擷取畫面只能有一個執行個體。) |
"shots":[
{
"id":0,
"keyFrames":[
{
"id":0,
"instances":[
{
"thumbnailId":"00000000-0000-0000-0000-000000000000",
"start":"0:00:00.209",
"end":"0:00:00.251",
"duration":"0:00:00.042"
}
]
},
{
"id":1,
"instances":[
{
"thumbnailId":"00000000-0000-0000-0000-000000000000",
"start":"0:00:04.755",
"end":"0:00:04.797",
"duration":"0:00:00.042"
}
]
}
],
"instances":[
{
"start":"0:00:00",
"end":"0:00:06.34",
"duration":"0:00:06.34"
}
]
},
]
brands
Azure AI Video Indexer 會在語音轉換文字記錄和/或影片 OCR 中偵測公司和產品名稱。 此資訊不包括品牌或標誌偵測的影像辨識。
名稱 | 描述 |
---|---|
id |
品牌的識別碼。 |
name |
品牌的名稱。 |
referenceId |
品牌 Wikipedia URL 的尾碼。 例如,Target_Corporation 是 https://en.wikipedia.org/wiki/Target_Corporation 的尾碼。 |
referenceUrl |
品牌的 Wikipedia URL (如果存在的話)。 例如: https://en.wikipedia.org/wiki/Target_Corporation 。 |
description |
品牌的描述。 |
tags |
與此品牌相關聯的預先定義標記清單。 |
confidence |
Azure AI Video Indexer 品牌偵測器的信賴值 (0 -1 )。 |
instances |
此品牌的時間範圍清單。 每個執行個體都有一個 brandType 值,指出此品牌曾出現在文字記錄還是 OCR 中。 |
"brands": [
{
"id": 0,
"name": "MicrosoftExcel",
"referenceId": "Microsoft_Excel",
"referenceUrl": "http: //en.wikipedia.org/wiki/Microsoft_Excel",
"referenceType": "Wiki",
"description": "Microsoft Excel is a sprea..",
"tags": [],
"confidence": 0.975,
"instances": [
{
"brandType": "Transcript",
"start": "00: 00: 31.3000000",
"end": "00: 00: 39.0600000"
}
]
},
{
"id": 1,
"name": "Microsoft",
"referenceId": "Microsoft",
"referenceUrl": "http: //en.wikipedia.org/wiki/Microsoft",
"description": "Microsoft Corporation is...",
"tags": [
"competitors",
"technology"
],
"confidence": 1.0,
"instances": [
{
"brandType": "Transcript",
"start": "00: 01: 44",
"end": "00: 01: 45.3670000"
},
{
"brandType": "Ocr",
"start": "00: 01: 54",
"end": "00: 02: 45.3670000"
}
]
}
]
統計資料
名稱 | 描述 |
---|---|
CorrespondenceCount |
影片中的信件數目。 |
SpeakerWordCount |
每個說話者的字數。 |
SpeakerNumberOfFragments |
說話者在影片中具有的片段數目。 |
SpeakerLongestMonolog |
說話者最長的獨白。 如果說話者在獨白中有無聲的部分,也會包含在其中。 獨白開頭和結尾的無聲部分則會被移除。 |
SpeakerTalkToListenRatio |
將說話者獨白的時間 (不含無聲的部分) 除以影片的總時間長度。 時間會四捨五入至小數點第三位。 |
audioEffects (預覽)
名稱 | 描述 |
---|---|
id |
音訊效果的識別碼。 |
type |
音訊效果的類型。 |
name |
語言中音訊效果的類型,而 JSON 已在該語言中編製索引。 |
instances |
此音訊效果曾出現的時間範圍清單。 每個執行個體都有一個信賴度欄位。 |
start + end |
原始影片中的時間範圍。 |
adjustedStart + adjustedEnd |
時間範圍與調整後的時間範圍。 |
audioEffects: [{
{
id: 0,
type: "Laughter",
name: "Laughter",
instances: [{
confidence: 0.8815,
adjustedStart: "0:00:10.2",
adjustedEnd: "0:00:11.2",
start: "0:00:10.2",
end: "0:00:11.2"
}, {
confidence: 0.8554,
adjustedStart: "0:00:48.26",
adjustedEnd: "0:00:49.56",
start: "0:00:48.26",
end: "0:00:49.56"
}, {
confidence: 0.8492,
adjustedStart: "0:00:59.66",
adjustedEnd: "0:01:00.66",
start: "0:00:59.66",
end: "0:01:00.66"
}
]
}
],
人氣
情緒會依其 sentimentType
欄位 (Positive
、Neutral
或 Negative
) 進行彙總。 例如:0-0.1
, 0.1-0.2
。
名稱 | 描述 |
---|---|
id |
情緒的識別碼。 |
averageScore |
所有該情緒類型執行個體的總分平均值。 |
instances |
此人氣曾出現的時間範圍清單。 |
sentimentType |
類型可以是 Positive 、Neutral 或 Negative 。 |
"sentiments": [
{
"id": 0,
"averageScore": 0.87,
"sentimentType": "Positive",
"instances": [
{
"start": "00:00:23",
"end": "00:00:41"
}
]
}, {
"id": 1,
"averageScore": 0.11,
"sentimentType": "Positive",
"instances": [
{
"start": "00:00:13",
"end": "00:00:21"
}
]
}
]
visualContentModeration
visualContentModeration
文字記錄包含 Azure AI Video Indexer 發現可能含有成人內容的時間範圍。 如果 visualContentModeration
是空的,則不會識別任何成人內容。
含有成人或猥褻內容的影片只能供私人觀看。 使用者可以提交人工檢閱內容的要求。 在此情況下,IsAdult
屬性會包含人工檢閱的結果。
名稱 | 描述 |
---|---|
id |
視覺內容仲裁的識別碼。 |
adultScore |
成人分數 (由內容仲裁提供)。 |
racyScore |
辛辣分數 (由內容仲裁提供)。 |
instances |
視覺內容仲裁出現的時間範圍清單。 |
深入了解有關 visualContentModeration
"visualContentModeration": [
{
"id": 0,
"adultScore": 0.00069,
"racyScore": 0.91129,
"instances": [
{
"start": "00:00:25.4840000",
"end": "00:00:25.5260000"
}
]
},
{
"id": 1,
"adultScore": 0.99231,
"racyScore": 0.99912,
"instances": [
{
"start": "00:00:35.5360000",
"end": "00:00:35.5780000"
}
]
}
]
textualContentModeration
名稱 | 描述 |
---|---|
id |
文字內容仲裁的識別碼。 |
bannedWordsCount |
禁用文字數目。 |
bannedWordsRatio |
禁用字與總字數的比例。 |
深入了解 textualContentModeration
emotions
Azure AI Video Indexer 會根據語音和音訊提示來識別情緒。
名稱 | 描述 |
---|---|
id |
表情的識別碼。 |
type |
所識別表情的類型:Joy 、Sadness 、Anger 或 Fear 。 |
instances |
這一個表情出現的時間範圍清單。 |
"emotions": [{
"id": 0,
"type": "Fear",
"instances": [{
"adjustedStart": "0:00:39.47",
"adjustedEnd": "0:00:45.56",
"start": "0:00:39.47",
"end": "0:00:45.56"
},
{
"adjustedStart": "0:07:19.57",
"adjustedEnd": "0:07:23.25",
"start": "0:07:19.57",
"end": "0:07:23.25"
}]
},
{
"id": 1,
"type": "Anger",
"instances": [{
"adjustedStart": "0:03:55.99",
"adjustedEnd": "0:04:05.06",
"start": "0:03:55.99",
"end": "0:04:05.06"
},
{
"adjustedStart": "0:04:56.5",
"adjustedEnd": "0:05:04.35",
"start": "0:04:56.5",
"end": "0:05:04.35"
}]
},
{
"id": 2,
"type": "Joy",
"instances": [{
"adjustedStart": "0:12:23.68",
"adjustedEnd": "0:12:34.76",
"start": "0:12:23.68",
"end": "0:12:34.76"
},
{
"adjustedStart": "0:12:46.73",
"adjustedEnd": "0:12:52.8",
"start": "0:12:46.73",
"end": "0:12:52.8"
},
{
"adjustedStart": "0:30:11.29",
"adjustedEnd": "0:30:16.43",
"start": "0:30:11.29",
"end": "0:30:16.43"
},
{
"adjustedStart": "0:41:37.23",
"adjustedEnd": "0:41:39.85",
"start": "0:41:37.23",
"end": "0:41:39.85"
}]
},
{
"id": 3,
"type": "Sad",
"instances": [{
"adjustedStart": "0:13:38.67",
"adjustedEnd": "0:13:41.3",
"start": "0:13:38.67",
"end": "0:13:41.3"
},
{
"adjustedStart": "0:28:08.88",
"adjustedEnd": "0:28:18.16",
"start": "0:28:08.88",
"end": "0:28:18.16"
}]
}
],
主題
Azure AI Video Indexer 會從文字記錄中推斷主要主題。 如果可能,會包含第二層級的 IPTC 分類。
名稱 | 描述 |
---|---|
id |
主題的識別碼。 |
name |
主題的名稱。 例如: Pharmaceuticals 。 |
referenceId |
反映主題階層的階層連結。 例如: HEALTH AND WELLBEING/MEDICINE AND HEALTHCARE/PHARMACEUTICALS 。 |
confidence |
範圍內 0 -1 中信賴分數。 值越高,信賴程度就越高。 |
language |
主題中使用的語言。 |
iptcName |
IPTC 媒體程式碼名稱 (如果偵測到)。 |
instances |
目前,Azure AI Video Indexer 不會將主題編製為時間間隔索引。 整個影片會用作間隔。 |
"topics": [{
"id": 0,
"name": "INTERNATIONAL RELATIONS",
"referenceId": "POLITICS AND GOVERNMENT/FOREIGN POLICY/INTERNATIONAL RELATIONS",
"referenceType": "VideoIndexer",
"confidence": 1,
"language": "en-US",
"instances": [{
"adjustedStart": "0:00:00",
"adjustedEnd": "0:03:36.25",
"start": "0:00:00",
"end": "0:03:36.25"
}]
}, {
"id": 1,
"name": "Politics and Government",
"referenceType": "VideoIndexer",
"iptcName": "Politics",
"confidence": 0.9041,
"language": "en-US",
"instances": [{
"adjustedStart": "0:00:00",
"adjustedEnd": "0:03:36.25",
"start": "0:00:00",
"end": "0:03:36.25"
}]
}]
. . .
說話者
名稱 | 描述 |
---|---|
id |
說話者的識別碼。 |
name |
說話者的名稱,格式為 Speaker #<number> 。 例如: Speaker #1 。 |
instances |
此說話者曾出現的時間範圍清單。 |
"speakers":[
{
"id":1,
"name":"Speaker #1",
"instances":[
{
"adjustedStart":"0:00:10.21",
"adjustedEnd":"0:00:12.81",
"start":"0:00:10.21",
"end":"0:00:12.81"
}
]
},
{
"id":2,
"name":"Speaker #2",
"instances":[
{
"adjustedStart":"0:00:12.81",
"adjustedEnd":"0:00:17.03",
"start":"0:00:12.81",
"end":"0:00:17.03"
}
]
},
意見反應
https://aka.ms/ContentUserFeedback。
即將登場:在 2024 年,我們將逐步淘汰 GitHub 問題作為內容的意見反應機制,並將它取代為新的意見反應系統。 如需詳細資訊,請參閱:提交並檢視相關的意見反應