OCR の設定

以下の表に、AI Center で ML パッケージを作成する際の OCR の設定の詳細を示します。

注:

An upcoming deprecation is announced for the UseServerSideOCR option. Here you can find more details.

以下の表に、AI Center で ML パッケージを作成する際の OCR の設定の詳細を示します。

OCR ML パッケージ

ML パッケージ	詳細
OCR for Chinese, Japanese, Korean	None. Please leave OCR configuration empty when creating this package.
UiPathDocument OCR (UiPath ドキュメント OCR)	None. Please leave OCR configuration empty when creating this package.

再トレーニング不可の ML パッケージ

ML パッケージ	詳細
DocumentClassifier (ドキュメント分類)	None. Please leave OCR configuration empty when creating this package.
DocumentUnderstanding (ドキュメントの理解)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.

Out-of-the-box Pre-trained ML Packages

ML パッケージ	詳細
ACORD25 (賠償責任保険証明書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
ACORD125 (企業向け保険契約申込書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
ACORD126 (企業総合賠償責任保険)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
ACORD131 (アンブレラ/エクセス保険)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
ACORD140 (商業保険申込書の財物補償条項)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Bank Statements (銀行預金残高証明書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
BillsOfLading (船荷証券)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Checks (小切手)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Financial Statements (財務諸表)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
FM1003 (米国の統一住宅ローン申請書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
IDCards (ID カード)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Invoices (請求書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
InvoicesAustralia (請求書 - オーストラリア)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
InvoicesChina (請求書 - 中国)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. Only Google Cloud Vision OCR is currently supported. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
InvoicesIndia (請求書 - インド)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
InvoicesJapan (請求書 - 日本)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. Only Google Cloud Vision OCR is currently supported. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
I9 (米国の就労資格証明書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Packing Lists (梱包明細書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Passports (パスポート)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
PurchaseOrders (発注書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Receipts (領収書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
RemittanceAdvices (送金通知書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
UtilityBills (公共料金の請求書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
Vehicle Titles (自動車の権利書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
W2 (米国の源泉徴収票)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
W9 (米国の納税申告書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
4506T (米国の納税申告証明依頼書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.
990 (米国の所得税非課税団体申告書)	Optional. This OCR configuration is used when you check the UseServerSideOCR checkbox on the Machine Learning Extractor activity. This is also necessary for using the eval.redo_ocr environment variable in Evaluation Pipelines.

その他のすぐに使える ML パッケージ

ML パッケージ	詳細
Form Extractor (フォーム抽出器)	None. Please leave OCR configuration empty when creating package.
Intelligent Form Extractor (インテリジェント フォーム抽出器)	Required. Please fill in the URL of the HandwritingRecognition ML Skill which you need to deploy before creating this package.
Intelligent Keyword Classifier (インテリジェント キーワード分類器)	None. Please leave OCR configuration empty when creating this package.
HandwritingRecognition (手書き文字認識)	None. Please leave OCR configuration empty when creating this package.