Migration

.NET Standard SDK v2.0.0 Release Migration Guide:

This guide will help you migrate from v1.3.1 to v2.0.0 of the .NET Standard SDK by highlighting the major changes.

v2.0.0 of the .NET Standard SDK is migrated to Visual Studio 2017 and the source project can be used in Visual Studio Mac.

Each service of the SDK is programmatically generated.

First-time generated services include Speech to Text, Text to Speech, Visual Recognition and Natural Language Classifier.

Version date constants have been removed.

Speech to Text

• GetModels() is replaced with ListModels().
• ListCustomWords() is replaced by ListWords().
• wordType is a string instead of an enum.
• DeleteCustomWord() is replaced by DeleteWord().
• AddCustomWords() is replaced by AddWords().
• AddCustomWord() is replaced by AddWord().
• GetClassifiersPerClassifierVerbose is replaced by classifier.
• SpeechRecognitionEvent is replaced by SpeechRecognitionResults.
• WordDefinition is replaced by Word.
• CorporaProperty is replaced by _Corpora.
• CreateCustomModel() is replaced by CreateLanguageModel().
• DeleteCustomModel() is replaced by DeleteLanguageModel().
• DeleteCustomWord() is replaced by DeleteWord().
• TrainCustomModel() is replaced by TrainLanguageModel().
• ResetCustomModel() is replaced by ResetLanguageModel().
• UpgradeCustomModel() is replaced by UpgradeLanguageModel().
• ListCustomModel() is replaced by GetLanguageModel().
• ListCustomModels() is replaced by ListLanguageModels().
• ListCustomWord() is replaced by GetWord().
• ListCustomWords() is replaced by GetWords().
• AddCorpus() accepts a stream instead of a byte[].

Visual Recognition

• CreateClassifier now takes a single CreateClassifier object.
• UpdateClassifier now takes a single UpdateClassifier object.
• status is now an enum instead of a string.
• ClassName is replaced by _Class.
• GetClassifiersBrief() is replaced by ListClassifiers(verbose: false).
• GetClassifiersVerbose() is replaced by ListClassifiers(verbose: true).
• Classify() accepts a stream instead of a byte[].
• DetectFaces() accepts a stream instead of a byte[].