Uploaded image for project: 'Nuxeo AI Core'
  1. Nuxeo AI Core
  2. AICORE-619

AWS Transcribe results is not decoded correctly

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Open
    • Priority: Critical
    • Resolution: Unresolved
    • Affects Version/s: 3.5.10
    • Fix Version/s: None
    • Component/s: AI Cloud Services

      Description

      The transcribe result downloaded from s3 is not decoded correctly because s3 returns application/octet-stream as the mimetype so the text is decoded as ascii characters instead of UTF8.

      https://github.com/nuxeo/nuxeo-ai/blob/bdce9a26be1859694c327902d7d0f1e609d83c78/addons/nuxeo-ai-aws-core/src/main/java/org/nuxeo/ai/enrichment/async/TranscribeEnrichmentProvider.java#L158

        Attachments

        1. Screenshot 2023-05-26 112907.png
          166 kB
          Michaël Vachette
        2. Screenshot 2023-05-26 113214.png
          127 kB
          Michaël Vachette
        3. Screenshot 2023-05-26 113522.png
          39 kB
          Michaël Vachette

          Issue Links

          links to

          Web Link Github PR

            Activity

              People

              • Votes:
                0 Vote for this issue
                Watchers:
                2 Start watching this issue

                Dates

                • Created:
                  Updated: