[BUG] Error using vector search with ImageVectorFieldNames

[druidajes]

Library name and version

Azure.AI.OpenAI 2.0.0.0

Describe the bug

When I use vector search with ImageVectorFieldNames, an error occurs

ChatCompletionOptions options = new();

options.AddDataSource(new AzureSearchChatDataSource()
{
    Endpoint = _optionsAzureAISearch.Endpoint,
    IndexName = _optionsAzureAISearch.IndexName,
    Authentication = DataSourceAuthentication.FromApiKey(_optionsAzureAISearch.Key),
    QueryType = DataSourceQueryType.VectorSemanticHybrid,
    SemanticConfiguration = _optionsAzureAISearch.IndexName + "-semantic-configuration",
    VectorizationSource =  DataSourceVectorizer.FromDeploymentName(_optionsAzureOpenAIOptions.EmbeddingsModelDeploymentName),
    InScope = true,
    TopNDocuments = 3,
    FieldMappings = new DataSourceFieldMappings()
    {
        ContentFieldNames = { "chunk" },
        FilePathFieldName = "metadata_storage_name",
        TitleFieldName = "title",
        UrlFieldName = "metadata_storage_path",
        VectorFieldNames = { "text_vector" },
        ImageVectorFieldNames = { "image_vector" }
    },

});

However, if I don't use ImageVectorFieldNames, no error occurs.

Expected behavior

That you can search for image vectors

Actual behavior

Cannot search for image vectors

Reproduction Steps

The exception is raised from the code provided above

Environment

No response

[github-actions[bot]]

Thank you for your feedback. Tagging and routing to the team member best able to assist.

[AngelosP]

Hi @druidajes ,

Thank you for raising this issue! Please hang tight while we take a look and feel free to add more details if you have any.